WO2015178377A1

WO2015178377A1 - Assembly instruction device and method

Info

Publication number: WO2015178377A1
Application number: PCT/JP2015/064328
Authority: WO
Inventors: 博文田口; 敦子榎本; 中須　信昭; 俊一川邊; 裕美子上野; 典明山本
Original assignee: 株式会社日立製作所
Priority date: 2014-05-20
Filing date: 2015-05-19
Publication date: 2015-11-26
Also published as: JP6425718B2; JPWO2015178377A1

Abstract

Provided are an assembly instruction device and method that automate the teaching of an appropriate hand selection task while taking into consideration compatibility with the component to be grasped. This assembly instruction device selects a hand for grasping a component in the assembly task for an assembled product, and is provided with: a storage unit that stores a processing program (hand selection processing program) that selects a hand, and a relational database; and a data processing unit that executes the hand selection processing program. The relational database includes: a component information database that stores information pertaining to components; and a hand information database that stores information pertaining to one or more hands. By executing the hand selection processing program, the data processing unit generates component grasping position information or generates teaching information for selecting the hand that grasps the component on the basis of the information in the component information database and hand information database.

Description

Assembly teaching apparatus and method

The present invention relates to a part assembly teaching apparatus and method.

When generating assembly operation programs in advance in assembly automation using robots, direct teaching is generally performed while operating the robot directly. In this direct teaching, it is necessary to teach a series of motions desired for the robot for each operation, and the teaching work becomes complicated, which requires a great amount of man-hours, which is a heavy burden on the teaching worker.

In recent years, due to the diversification of customer needs, the production form of products has shifted from a low-mix, high-volume production type to a multi-variable, variable-volume production type, and the assembly process using robots is also expected to support multiple types. ing.
This increases the burden on the operator who performs direct teaching at an accelerated rate.

Therefore, conventionally, an automated technique for reducing the work burden of teaching has been proposed. For example, in Patent Document 1, an automatic tool changer is simplified by recognizing identification parts and grip positions of a plurality of tools by an image processing apparatus, automatically selecting and holding a tool, and automatically changing the tool. A technique for reducing the work burden of teaching related to selection is disclosed.

Further, for example, Patent Document 2 discloses a technique for automating teaching of a component gripping operation by a robot hand.

JP-A-62-264839 JP 2008-272886 A

In the case of an assembly line using a robot that has several types of general-purpose hands, when trying to support the assembly of new parts, an appropriate hand is selected and selected from existing hands, and the gripping position of the parts is further determined. It is necessary to decide. In this hand selection and gripping position determination, it is necessary to select a hand that best suits the characteristics of the part in consideration of the size, weight (mass), frictional force, etc. of the part.

However, the teaching work for selecting the hand is particularly complicated in the teaching work as a whole, and is a heavy burden on the conventional worker. Therefore, it is highly necessary to automate the teaching work.

However, the prior art disclosed in Patent Document 1 selects one of a plurality of pre-given hands without considering any compatibility or compatibility with a part to be gripped by the hand. is there.

The prior art disclosed in Patent Document 2 determines the gripping position of a component when a predetermined hand grips the predetermined component, and does not consider compatibility or compatibility with the component in the selection. .

SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to provide an assembly teaching apparatus and method for automating teaching of an appropriate hand selection operation considering compatibility with a gripped component or optimizing the position of a gripped component. .

The present invention relates to an assembly teaching apparatus for selecting a hand that holds a part in an assembly operation of an assembly, a processing program for selecting a hand (hand selection processing program), a storage unit for storing a related database, and a hand selection process. A data processing unit that executes a program, and the related database includes a component information DB that stores information related to a component, and a hand information DB that stores information related to one or more hands. By executing a selection processing program, teaching information for selecting a hand that grips a component is generated based on information in the component information DB and the hand information DB, and component gripping position information is generated It can be realized as an assembly teaching apparatus characterized by this.

Further, the present invention can be realized as an assembly teaching method having a step of executing the above processing.

According to the present invention, the robot teaching work man-hour can be significantly reduced by automating the selection of the hand and the generation of the grip position information of the parts.

The block diagram of the assembly teaching apparatus concerning one Example of this invention is shown. It is a figure which shows hand tool information DB concerning one Example. It is a figure which shows components information DB concerning one Example. It is a figure which shows the whole processing flow of the assembly teaching apparatus concerning one Example. It is a figure which shows a part of processing flow of the assembly teaching apparatus concerning one Example. It is a figure which shows arrangement | positioning of the components of the assembly teaching apparatus concerning one Example. It is a figure which shows the 2 claw hand and its virtual hexahedron concerning one Example. It is a figure which shows the virtual hexahedron concerning one Example, and a larger target component. It is a figure which shows the virtual hexahedron concerning one Example, and a target object wider than it. It is a figure which shows the 3 claw hand and its virtual hexahedron concerning one Example. It is a figure which shows the movable range of 2 claw hand concerning one Example, and the width | variety of object component. An example of the input screen of the assembly teaching apparatus concerning one Example is shown. It is a figure which shows the outline of an example of the calculation method of the holding | grip position where the moment concerning one Example becomes the minimum. It is a figure and graph which show adsorption | suction of the components by the adsorption | suction hand concerning one Example. It is a figure and graph which show adsorption of parts by an adsorption hand with two adsorption surfaces concerning one example. It is a figure which shows the example of holding | grip of the ring-shaped components by the hand of the outward opening concerning one Example, and a graph. It is a figure which shows the hand which has 2 types of hand functions concerning one Example. It is a figure which shows the processing flow of component gripping surface extraction concerning one Example.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

Embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a configuration diagram of an assembly teaching apparatus 100 according to this embodiment.

The assembly teaching apparatus 100 includes a data processing unit (CPU 10) such as a computer that executes a program and processes data, an input unit 20 that receives an input signal including an instruction from a user, and a display unit 30 that displays at least an input screen. A storage unit 40 for storing parts and hand databases and processing programs, a hand 3D-CAD device 101 for generating hand design data, a product 3D-CAD device 102 for generating part design data, And a vision system 103 such as a camera or a scanner capable of capturing product shape data.

The storage unit 40 includes a parts information DB 104 that stores information on the shape, size, material, surface state, mass, and the like of parts obtained from the product 3D-CAD device 102 and the vision system 103, and a hand, a driver, a jig, and the like. A hand / tool information DB 117 that stores tool information and a hand selection program 50 that is a processing program for selecting an optimum hand are stored.

The hand selection program 50 includes an assembly order generation unit 105 that generates an assembly order and an assembly direction of a product composed of a plurality of parts, a component gripping surface extraction unit 106 that extracts a surface or position that can be gripped by the hand on the component, A part-hand information collating unit 107 that collates information on a hand and a tool, a part-hand compatibility evaluation unit 108 that evaluates compatibility between a part and a hand, a hand determination unit 109 that determines an optimum hand, A component gripping position generation unit 110 that generates a gripping position of the component when gripping the target component with the determined hand, and a path generation unit 111 that generates a path for moving each component from the initial position to the assembly position. The hand switching frequency counting unit 112 that measures the number of hand switching times (replacement times) until the assembly of the parts is completed, and the assembly time for calculating the assembly time of the parts A calculation unit 113, stores.
Here, the hand 3D-CAD device 101 and the product 3D-CAD device 102 may be configured as a single device having both functions. Among them, in the hand tool information DB 117, as an example of the contents of the hand tool information, as shown in FIG. 2 (a), as shown in FIG. Various types of shape information such as a type that grips a part by closing the inside, a type that grips a part by opening the outside, a type that handles the product by suction instead of a nail, a finger type, a hand opening / closing operation range, Various information such as the load capacity is stored.

In the assembly order generation unit 105 and the component gripping surface extraction unit 106, information obtained by exhaustively extracting the parts that can be gripped by the component gripping surface extraction unit 106, the assembly order and assembly generated by the assembly order generation unit 105 Depending on the direction information, the surface on which a part is placed during assembly or the surface that needs to be shared with other parts during assembly is not gripped (cannot be gripped) by the hand. Is provided (details will be described later).

In addition, the component-hand information collating unit 107 includes component information (see FIG. 2B) such as the shape, size, and mass of the component obtained from the component information DB 104 in order to hold the component with the hand, and the hand and tool. Check the information.

The part-hand information collating unit 107 compares the size of the part and the movable range of the hand, compares the mass of the part and the portable mass of the hand, and considers the degree of slippage between the hand and the part. Extract candidates. After that, for each hand candidate, the part-hand compatibility evaluation unit 108 qualitatively or quantitatively calculates and evaluates the compatibility between the part and the hand using, for example, an evaluation function.

In the component-hand information collating unit 107 and the component-hand compatibility evaluating unit 108, a plurality of types of hands prepared in advance so as to correspond to an arbitrary component, for example, a suction type that sucks and grips a component, A gripping type in which a part is sandwiched and gripped by a plurality of plate-like or claw-shaped members, and an open-open type in which a hollow type part is opened and gripped by opening a plurality of plate-like or claw-shaped members, Assess compatibility qualitatively or quantitatively. Here, details regarding the method of evaluating the combination of parts and hands according to the present embodiment will be described later.

Regarding the hand and tool evaluated by the part-hand compatibility evaluation unit 108, the hand and tool evaluated by the part-hand compatibility evaluation unit 108 for each part in the hand determination unit 109, which is the next processing, For example, the hand or tool having the maximum evaluation value is determined. Here, the evaluation value should be determined with the hand or tool having the maximum value or the hand or tool with the minimum value, depending on the setting of the evaluation function (described later) that is the basis for the calculation. Is different.

When the hand / tool is determined by the hand determination unit 109, the component gripping position generation unit 110 determines which region of the component should be gripped by the hand or tool (details will be described later). Subsequently, a path for moving each part from the initial position to the assembly position where the parts are assembled is generated by the path generation unit 111. Thereafter, the hand switching count (number of replacements) until the assembly of the parts is measured by the hand switching counting unit 112, and then the assembly time of the parts is calculated by the assembly time calculating unit 113.

The operation flow of the assembly teaching apparatus according to the present embodiment is shown in FIG. Here, the hand considered in this embodiment is a hand having two claws and three claws as an example, a type that grips a part by closing inside, a type that grips a part by opening outside, or a suction type, or A finger type (not shown) imitating a human hand, a hand configured by a combination of them, and the like are conceivable.

When the start of the assembly teaching apparatus is started (START S2000), the assembly sequence generation unit 105 generates the assembly procedure and the direction for assembling the parts based on the product information of the product 3D-CAD apparatus 102 (S2001). .

Next, the number of parts is set to N and the number of hands is set to n (S2002). Next, the values of the counter K that counts the number of parts and the counter Q that counts the number of hands in this operation flow are respectively initialized (value is set to 0) (S2003).

Subsequently, the counter K is incremented by 1 (S2004). First, based on the information of the part-hand information collating unit 107 in the hand selection program 50 in the storage unit 40 for the first part of assembly, that is, the first part. Hand candidates that can be held are narrowed down to the first part and allocation corresponding to the part is performed (S2005). Subsequently, the number of component gripping surfaces is confirmed (S2006). If the presence of two or more surfaces is confirmed, the process proceeds to the next step, and the hand count counter Q is incremented by 1 (S2008). Next, the process proceeds to the step of selecting each type of hand, for example, for each type of hand of nail type, suction type, and finger type (S2009). Here, since the suction-type hand can be sucked if at least one suction surface is present, it is possible to cope with conditions of two or more surfaces (S2009).

The component gripping surface extraction unit 106 extracts the component gripping surface of the first component (S2010), and generates a virtual hexahedron (hereinafter referred to as a hexahedral virtual space) serving as an index of a size that can be gripped by the hand (S2011). ). Details of this virtual hexahedron will be described later.

Subsequently, the virtual hexahedron space formed by the two-claw or three-claw type hand is compared with the size of the part to be gripped, and it is evaluated whether or not the first part can be gripped by the first hand ( S2012). Next, when gripping a component with the hand, a gripping position is calculated such that the moment is minimized in the operation after gripping (in a grippable plane) (S2013).

Here, an outline of an example of a method for calculating the gripping position at which the moment is minimized will be described with reference to FIG.

Here, for the sake of explanation, the moving direction of the hand is the + Z direction (220Z) in the figure.
In FIG. 12A, the component gripping center 220c and the component gripping center position 223g and the component gripping center of the component 223, the protrusion 223a on the component 223, the center of gravity position 223g of this component, the hand 220, and the two claws 222 of the hand The distance from 220c is L, the mass of the part is m, the gravitational acceleration is G, and the moment 225 generated around the gripping center 220c is a value obtained by multiplying the product 223f of the part mass m × gravity acceleration G by the distance L × M × G ×. L. Here, when the gripping center 220c can approach the center of gravity 223g of the component, it is desirable to reduce the moment m × G × L by bringing the hand close to the position of the center of gravity as close as possible.

Here, FIG. 12B is a diagram in which the hand 220 is closest to the center of gravity position 223g of the component in FIG. 12A, but since the component 223 has a protrusion 223a, L = 0. The hand 220 cannot be brought close to the part center of gravity 223g to the position. 12C, the horizontal axis indicates the distance L from the center of gravity of the component (223g) to the hand gripping center 220c, and the vertical axis indicates the moment applied to the hand gripping center 220c.

Further, the state of L = L2 is shown in FIG. 12 (a), and the state of L = L1 is shown in FIG. 12 (b). From the above, it is desirable to make L = 0 closer to the moment to be applied to the center of the gripping position. However, if this is difficult due to the gripped part, the structure and shape of the hand, In the example of c), L = L1 and moment MO = MOmin. The position where L is minimized is calculated by the component gripping position generation unit 110 based on the data of the product 3D-CAD device 102. Next, the assembly teaching flow will be described with reference to FIG. Hand candidates are extracted from a plurality of hands, and a gripping position for gripping a component with the hands is calculated (S2014).

Next, in the combination of the first part and the hand that holds it, a path from the part gripping to the part assembly is generated by the path generation unit 111, and there is a part, a hand, or a robot arm (not shown) in the path. Also, it is confirmed whether or not interference occurs with the surrounding environment of the robot, for example, an interference object such as a part supply table or a cage that forms the assembly table (S2016). Here, when there is no interference, as the next step, the cumulative total assembly time and / or the number of hand switching (exchange) are counted (S2017). Here, since it is necessary to perform the same process with a plurality of hands on one part and calculate the compatibility evaluation between the part and the hand, the loop with A is repeated n times for the number of hands.

Further, since the same repetition is required for N parts, which is the number of parts, the process is repeated with S2004. When the series of N × n loops is completed, a hand combination that minimizes the total component assembly time T is extracted (S2020), and an actual robot operation program is automatically generated (S2021).

Here, the hand combination that minimizes the total component assembly time T can be considered as a low-cost assembly operation.
By executing these processing flows, it is possible to determine a hand replacement schedule such as the type of hand to be replaced corresponding to the parts to be assembled, the number of hand replacements, and the hand replacement timing.

Subsequently, when the operation program generated in S2021 is executed (S2022), a hand corresponding to a part necessary for assembly is attached (S2023), a series of assembly operations for a product (not shown) is executed (S2024), and the operation ends. (S2025).

Subsequently, in FIG. 3, the combination evaluation method of the hand and parts described in S2012 will be described.

First, the magnitude relationship between the hand and the component is evaluated (S2012a), then the component mass and the load resistance of the hand (hereinafter referred to as mass resistance) are evaluated (S2012b), and the sliding condition between the component surface and the gripping surface of the hand is further evaluated. Evaluation is performed (S2012c), and the suitability of the hand with respect to an arbitrary part is calculated.

The hand suitability for this arbitrary part will be described in detail later with reference to FIG. Here, FIGS. 13 to 14 show an outline of the suction type hand.

The suction hand 230 having one suction surface shown in FIG. 13A is composed of a hand base 231, a suction hollow shaft 232, and a suction cup 233. 230c is a substantially center line of the

suction hand

230, 234 is a part to be assembled, 234g is a center of gravity of the part, and 234c is a center of gravity line passing through the center of gravity of the part. Further, the suction hand 230 shown in FIG. 13A moves in the −Z direction at the bottom of the drawing to try to suck the component 234. Here, the mass of the component 234 is m.

FIG. 13B is a diagram in the middle of the movement of the component 234 in the + Z direction above the drawing when the component 234 is picked up by the suction hand 230. FIG. 13C shows the position of the center of gravity 234g of the component 234 in FIG. 13A and the size (width in FIG. 13) of the component. The part is a substantially rectangular parallelepiped, the width of the part is 2L, and the distance from the center of gravity position to one end is L. Further, (d) in FIG. 13 in FIG. 13 (c) shows a state in which substantially the center line 230c of gravity line 234c and suction hand parts are shifted by _{L 3.}

Also, FIG. 13 (f) shows the relationship of moment M generated when a part is picked up and lifted by a suction hand. The horizontal axis is the distance L from the center of gravity of the component to the center of gripping (suction) of the suction hand, the vertical axis is the moment MO applied to the suction part, and the straight line 240 is a straight line of the moment MO and is a line of mass m × gravity acceleration G × distance L. Yes. Assuming that the mass m of the component does not change during the adsorption, it goes without saying that the distance L from the center of gravity of the component must be reduced in order to reduce the moment MO. In the example shown in FIG. 13, the component suction surface 234T is in a substantially flat state, so that suction with a suction hand is possible in an area within the component suction surface 234T.

Here, an example in which the part shapes are different will be described with reference to FIG. In FIG. 13E, parts other than the part related to the component 235 are common in FIG. The component 235 shows an example in which the surface on which the component is formed is not a uniform surface, and the suction surface 235T is partially convex. When the component 235 is to be sucked by the suction hand 230, the component gripping position generation unit 110 may preferentially calculate a surface other than the suction surface 235T as a preferential suction surface. In (e), the case where the suction surface 235T is used as the suction surface will be described for the sake of explanation.

As shown in FIG. 13E and as described above, the smaller the distance from the center of gravity of the part to the hand gripping position, the smaller the influence of the moment and the smaller the load on the hand. However, as shown in FIG. 13 (e), when there is a protrusion 235U that cannot be sucked by the suction hand near the center of gravity line 235c of the component, it is necessary to avoid that place at a minimum distance. As an example of this avoidance method, the avoidance method is the method described in FIG. This avoidance process is calculated by the component gripping surface extraction unit 106. Next, description will be given of a case where a component is picked up by another type of hand.

14A and 14B, the suction hand 240 having two suction surfaces includes a base 241 of the hand, suction hollow shafts 242l and 242r, and suction cups 243l and 243r.

Reference numerals

240a and 240b denote center lines of the suction hollow shafts 242l and 242r. Reference numeral 244 denotes a part to be assembled, and 244g denotes a position of the center of gravity. Similarly to FIG. 13A, the suction hand 240 shown in FIG. 14A moves in the −Z direction below the drawing, and tries to suck the suction surface 244T of the component 244. FIG. 14B is a diagram in which the component 244 is attracted to the suction hand 240 by the suction surface 244T and is moving in the + Z direction above the drawing.

(C) in FIG. 14 shows a moment diagram. From this figure, it can be seen that when the center of gravity position 234g is between the suction cups (between -i and + i), less moment is generated than when the center of gravity is not. This indicates that the load on the hand 240 is small. In this figure, these calculations are also performed by the component gripping position generation unit 110, and it is possible to reduce the load on the upper device to which the hand or a hand (not shown) is attached. As shown in FIG. 13 and FIG. 14, the suction type hand basically requires only one surface for picking up the components. Therefore, the handling of components having complicated shapes or components having a small plane area is limited. It is effective for.

4A shows an example of a flow applied to a suction type hand, and FIG. 4B shows an example of a flow applied to a finger type hand (not shown). In the case of the suction type hand shown in FIG. 4 (a), if two or more parts can be gripped by the hand in S2006 of FIG. 3, first, extraction of the part suction surface is attempted ((a of FIG. 4). ) S3001). Here, when the suction surface is not found, the processing flow moves to D in (b) of FIG. 4B in the case of using a finger type hand described later.

When the suction surface is found in S3001, the degree of compatibility between the hand and the part is calculated (S3002). Subsequently, when the hand sucks the part and the hand tries to move to the next posture, the hand suction position is calculated so that the moment acting on the hand is minimized by the mass of the part (S3003). The suction position is determined (S3004), and the operation temporarily moves to S2015 shown in FIG. Since the method of minimizing this moment in a grippable state has been described with reference to FIGS. 12 and 13, it is omitted here. Here, the collation between the suction type hand and the part in S3002 will be described next.

S3002a is a step of evaluating the mass of the part and the mass resistance of the hand. When the mass of the part obtained from the attribute information of the 3D-CAD information is likely to exceed the mass resistance of the hand, the combination of the part and the hand is inappropriate. And the process is moved to A shown in FIG. 3, and the hand is reviewed again. On the other hand, if the mass resistance is not exceeded in step S005, the sliding condition of the part and the hand is evaluated in the next S3002b.

Here, the information used as the basis for the evaluation of the slip condition is, for example, attribute information possessed by the 3D-

CAD apparatuses

101 and 102. Next, FIG. 4B is used to show a flow when a finger type hand (not shown) is selected.
Here, the finger-type hand is a hand provided with one or more pairs of finger-shaped connecting members having joints.

4 When the suction surface cannot be detected in S3001 of FIG. 4A, the flow proceeds to the flow using the finger type hand of FIG. 4B. First, in S3007, each step (S3007a and S3007b) of combination evaluation of a finger type hand and a component to be grasped is performed. Details of this evaluation will be described later. After this assembly evaluation, the moment applied to the hand when the part is gripped is calculated (S3008). For example, when the gripping position is arbitrarily changed at several places, the moment amount changes, but the moment amount is calculated at least at a plurality of grippable positions, and the gripping position is determined so as to be minimized (S3009).

If the component mass is heavier than the mass of the hand, the process moves to A (S2007) shown in FIG. 3 as the load is over, and the flow enters the flow for selecting the next hand candidate.
If the mass of the component is less than or equal to the withstand mass of the hand, the process proceeds to the step of evaluating the sliding condition of the component and the hand (S3007b).

Subsequently, extraction of the hand gripping position taking into consideration the product assembly procedure, which is a feature of the present embodiment, will be described with reference to FIG. FIG. 5A shows a workbench 180 for assembly, a part A181 constituting one part of the product, and a part B182. Parts A and B are parts that have not been assembled yet. A state of being placed on the work table 180 in a level state is shown.

Here, the surface 181a constituting one surface of the component 181 is a surface that comes into contact with the component B when assembled in the future, and the surface 182b is a surface facing the surface 182a. The parts 182 can be gripped by using the

surfaces

182a and 182b. A hatched surface 182a of the component B182 indicates a surface that contacts the component A181 when assembled in the future. FIG. 5B is a view in the process of assembling the part 181 and the part 182, and FIG. 5C shows the final form in which the part A181 and the part B182 are assembled. Here, 181a and 182a, which are the contact surfaces of the two parts after assembly, are sent to the assembly order generation unit 105 based on the product information DB 104 obtained by the product 3D-CAD device 102 or the vision system 103 in FIG. Is extracted.

Here, the hand to be used is the same for picking up parts and assembling parts. Further, consider a case where the parts are not changed during the transition from the part removal to the part assembly.

In the part B182 of FIG. 5A, the surface 182c is in contact with the workbench 180. For example, the

surfaces

182a and 182b can be gripped by using a hand other than the surface 182c. However, when the state after the completion of assembly shown in FIG. 5C is predicted, the surface 182a of the part B and the surface 181a of the part A are in contact with each other. It cannot be. Therefore, at the used hand determined by the hand determining unit 109 and the gripping position determined by the component gripping position generating unit 110, the state after completion of assembly is calculated in advance by the assembly order generating unit 105, and the gripping or suctioning with the hand is performed. In this case, the gripping surface is determined by excluding the contact surface after assembly in advance. This process is a major feature of the present embodiment, and FIG. 17 shows a diagram representing these in the processing flow.

Further, the flow shown in FIG. 17 is performed by the component gripping surface extraction unit 106 in the component gripping surface extraction step (S2010 in FIG. 3) (details omitted). Next, an example of a method for determining whether or not an arbitrary part can be held with an arbitrary hand will be described with reference to FIGS. The contents described here are processed in the part-hand information collating unit 107 in FIGS. 1 and 3 and the step S2012a of the hand-part magnitude relation evaluation step.

FIG. 6 is a diagram schematically showing a two-claw hand. FIG. 6A is a schematic perspective view of a two-claw hand. The hand 190 includes a base portion 191 of the hand and two claws 192. These two claws open and close inward and outward in the X direction. By doing so, it is possible to grip the component. (In this description, the description of the actuator such as an actuator for driving the tab of the hand is omitted.) FIG. 6B is a diagram when the distance between the claw portions of the two-claw hand is maximized to Bmax. A state where a hexahedral virtual space 193 is formed in a region surrounded by the base portion 191 and two claws 192 is shown.

FIG. 6 (c) is a view of the two-claw hand as viewed from the front. The inter-nail dimension when the two nails are fully opened is Bmax, and the distance between the nails when the interval is closed to the minimum (dashed line 192a). The dimension is Bmin. FIG. 6D is a diagram in which the hexahedral virtual space 193 is extracted.
FIG. 7 shows a hexahedral virtual space 193 and a part 194 to be gripped by the two-claw hand.

FIG. 7A is a diagram showing the dimensional relationship between the hexahedral virtual space 193 and the part 194, and the width direction dimension Bp of the part satisfies the condition that Bmax> Bp with respect to the opening / closing direction dimension Bmax of the hand. It is a figure when satisfy | filling. FIG. 7B is a diagram in which the hexahedral virtual space 193 and the part 194 are overlapped, and since Bmax> Bp, the hexahedral virtual space formed by the base portion of the hand and two claws is illustrated. The state where the part 194 is included in the space 193, that is, the part 194 can be gripped by the two-claw hand 190 is shown.

In FIG. 7C, the state of the hexahedral virtual space 193 is the same as FIG. 7A and FIG. 7B, but the dimensional relationship of the part 195 to be grasped is different from that of the part 194. Shows the case. In FIG. 7C, Bmax <Bp. FIG. 7D is a diagram in which a hexahedral virtual space 193 and a component 195 are overlapped as in FIG. 7B. Here, as shown in FIG. 7D, since Bmax <Bp, the component 195 protrudes from the hexahedral virtual space 193. That is, the component 195 cannot be gripped by the two-claw hand 190.

In this way, the part-hand information collating unit 107 performs a process of superimposing the hexahedral virtual space that can be formed between the claws of the hand and the part to be grasped on the part-hand information collation unit 107, so that the two-nail type hand is applied to the part. It is possible to determine whether or not can be gripped. Next, another gripping example will be described with reference to FIG.

FIG. 8 (a) shows a hexahedral virtual space 193 and a part 194 formed by the two-claw hand shown in FIGS. Similarly to the above, 193a indicates a surface sandwiched between two claws.

In the state of FIG. 8A, since Bmax <Bp, the component 194 cannot be gripped by the two-claw hand 190 unless the relationship between the orientation of the two-claw hand and the component changes. FIG. 8B shows a state in which the direction of the two-claw hand 190 is rotated 90 degrees with respect to the component in FIG. In this state, the width of the hexahedron virtual space 193 with respect to the part is Bmax, and the part width is Wp. As shown in FIG. 8B, Bmax> Wp, and therefore, as shown in FIG. When the two are superposed on each other, the part 194 is included in the hexahedral virtual space 193.

In other words, if the surface direction gripped by the hand with respect to the part is fixed, and if the part is larger than the width of the hexahedral virtual space, the hand cannot be gripped, but the hand is intentionally rotated approximately 90 degrees. In this case, the component-hand information collating unit 107 in FIG. 1 performs collation assuming that the hand is rotated with respect to the component.

Subsequently, FIG. 9 is a diagram when the hand form is changed from the conventional two-claw hand to the three-claw hand. 9A is a schematic perspective view of the three-claw hand 200, FIG. 9B is a diagram in the case of assuming a hexahedral virtual space inside the claw as an example in FIG. 9A, FIG. (C) is a view of the three-claw hand 200 as viewed from above. The three-claw hand 200 is composed of a base portion 201 of the hand and three claws 202, and these three claws open and close inward and outward in the direction of arrow E in FIG. .

3 claw 202 is a view showing a state in which the claw interval is widened to the maximum, and 3 claws indicated by a broken line 202b are showing a state in which the component 204 is being held.
Since the relation between the virtual space of the hexahedron and the parts to be gripped is basically the same as the contents described with reference to FIGS. 6 to 8, the description thereof is omitted here.

Next, with reference to FIG. 10, an evaluation method related to the gripping operation of the hand and the component will be described using an example. 1 and 3, the part-hand compatibility evaluation unit 108 in FIG. 1 and FIG. 3 performs a hand / part magnitude relationship evaluation step S2012a, a part mass / hand mass resistance evaluation step S2012b, and a part / hand. The process is performed in the slip condition evaluation step S2012c.

FIG. 10 is a diagram showing a hand having a two-claw type configuration and parts to be gripped for simplicity.
Here, the two-claw hand 210 is composed of a base part 211 of the two-claw hand, a support shaft 211a of the hand, and two claws 212, and the state of the claw when the distance between the two claws is maximized is indicated by 212a. Similarly, the state of the nail when it is minimized is indicated by 212b.

The interval when the distance between the two claws is the maximum is indicated by L1, and the interval when the distance between the claws is the minimum is indicated by L2. Further, the width dimension of the component to be gripped here is indicated by Lp. Here, the ideal state desirable as a hand is that the ratio of Lp and L1 is sufficiently small when the nail is open ((Lp / L1) << 1), and the ratio of Lp and L2 is when the nail is closed. Is sufficiently large ((Lp / L2) >> 1), and if these conditions are satisfied, the opening / closing margin of the nail with respect to the size of the component increases. Therefore, as an example, the expression (1) is defined as an index of the opening / closing margin of the nail.
Lmargin (%) = [{1- (Lp / L1)} × 100 + {1-1 / (Lp / L2)} × 100]
= [{1- (Lp / L1)} + {1- (Lp / L2)}] × 100 ----- (1)
Also, when evaluating from the viewpoint of the mass of the component and the mass resistance of the hand, if the mass of the component is Mp and the mass mass of the hand is Mlimit, the smaller the mass of the component, the higher the margin. Become. Here, when the margin Mmargin with respect to mass is defined as an example, the equation shown in (2) is obtained.

Mmargin (%) = {1- (Mp / Mlimit)} × 100 ----- (2)
Further, the evaluation of the sliding condition between the part and the hand depends on the friction coefficient μ between the part and the hand and the force F with which the hand grips the part. The relationship between the part mass Mp, the friction coefficient μ, the gravitational acceleration G, and the force F with which the hand grips the part is balanced when the part is lifted in the vertical direction under the condition shown in the equation (3). It becomes a state.

μ × F = Mp × G ----- (3)
Further, it is possible to lift the part with the hand under the condition represented by the expression (4).

μ × F> Mp × G
μ> (Mp × G) / F (where μ ≦ 1) ----- (4)
Furthermore, if the moment applied to the hand is MOp and the moment allowable to the hand is MOlimit as an index of the margin of the moment applied to the hand, the margin is greater when the moment MOp applied to the hand is smaller than the allowable moment of the hand. Get higher. Here, when the margin MOmargin with respect to the moment is defined as an example, MOmargin (%) = {1− (MOp / Mlimit)} × 100 (5)
Can be defined as

As described above, when the evaluation index (evaluation function f) between the part and the hand in consideration of the size, mass, moment applied to the hand, and friction state is summarized, an expression (6) is given as an example. Here, the coefficients a, b, c, and d indicate weighting factors relating to the magnitude, mass, moment, and friction state, respectively. Needless to say, when a = b = c = d = 1, the magnitude, mass, and moment This indicates that the weight of the frictional state, that is, the importance is equal.

f = a × Lmargin + b × Mmargin + c × MOmargin + d × μ (6)
a, b, c, d: weighting factors
Further, the evaluation index may extend the equation of (6) and take the form of the root mean square of the size, mass, moment, and friction state as shown in the equation of (7).

f = {a × Lmargin ² + b × Mmargin ² + c × MOmargin ² + d × μ ² } ^1/2 ----- (7)
a, b, c, d: Weighting coefficient Next, FIG. 11 shows an example of an input screen of the assembly teaching apparatus in the present embodiment. This screen is displayed on the display unit 30 in FIG. 1, and an input operation is performed with the input unit 20 including a keyboard and a mouse. In FIG. 11, an input unit 501 for inputting a product name (file name) of data of the product 3D-CAD device 102, a display unit 502a for displaying 3D-CAD information of the product completion stored in the device 102, confirmation For example, a part layout layout is displayed in 502b.

Also, here, the hand gripping position surface designation menu 503 is selected so that the component gripping position in the hand can be input or designated, and the gripping position surface is designated by the

cursors

506a and 506b for hand gripping position surface designation in the display unit 504. Specify. In FIG. 11, 507a indicated by hatching and the opposite surface 508a of the nail on the opposite side are designated as the component gripping position surface of the hand. Here, the component gripping position surface of the hand may be data stored in the hand 3D-CAD device, and for example, the hand gripping surface of the claw may be set in advance by attribute setting or the like.
Reference numeral 509 denotes a confirmation screen for hand and component evaluation functions, 510 denotes an assembly sequence generation button, and 511 denotes an execution start button for the assembly teaching process shown in FIG.

In addition, after the CAD drawing information data has been fetched and the surface of the hand gripping position has been designated, a series of assembly teaching is automatically performed by pressing the teaching execution button.
Here, an example of a structure in which a hand has only one set of gripping portions has been described so far, but the scope of application of this embodiment is not limited to this. For example, as shown in FIG. It may have a structure having a gripping function, two or more sets of suction functions, or a gripping function and a suction function.

Here, an example in which the form of the hand is different will be described with reference to FIG.
A hand 250 shown in FIG. 15 includes a hand base 251 and three claws 252. Reference numeral 250 c denotes a center line of the hand 250. Reference numeral 253 denotes a ring-shaped component, and the three claws 252 have a structure suitable for gripping the component 253 at the center hole 235a portion.
As shown in FIG. 15, the three claws 252 are close to each other when the part 253 is passed, and after passing the part 253, the claws spread in the J direction in the figure to grip the part 253. can do.

Unlike the hand described so far, this hand has a form in which a part is gripped by opening outward.
FIG. 15 (f) shows the relationship between the movable range of the hand claw and the inner peripheral diameter of the ring-shaped part. Here, assuming that the movable range of the claw is R1 to R2 and the inner peripheral diameter of the ring-shaped part is rp1 to rp2, gripping is possible when R1 <rp1 <rp2 <R2. The calculation of the grippable range is also performed by the component-hand information matching unit 107 component. The method of selecting the hand, the method of determining the gripping position, and the concept of the opening / closing allowance of the parts and the hand are the same as the methods described so far, so detailed description thereof is omitted here.

FIG. 16 shows an example in which two hand functions are integrated into one hand.
FIG. 16 is a schematic perspective view of a hand 260 having a structure in which two systems of two-claw type hands are formed by one hand.

The hand 260 includes a base portion 261 of the hand and two sets of claws 262 (first claws) and 263 (second claws). The two sets of claws are independent on the inside and outside in the X direction. It opens and closes. In the drawing, 262a and 263a indicated by hatching indicate gripping surfaces for the parts in the

claws

262 and 263, respectively. The present embodiment can also be applied to a hand having a plurality of claws and suction cups for these one hand.

As described above, according to the present embodiment, even when a product is assembled using a new part, a hand (including a tool) suitable for handling the part is automatically selected, and the gripping position of the part is also selected. Generate automatically. In addition, since the hand selection information and the grip position information are reflected in the robot operation program, the teaching work can be made more efficient and labor-saving.

10 ... CPU (data processing unit)
20 ... input unit, 30 ... display unit, 40 ... storage unit,
50 ... Hand selection program, 100 ... Assembly teaching device,
104 ... parts information DB, 105 ... assembly order generation unit,
106: Component gripping surface extraction unit, 107: Component-hand information verification unit,
108: Component-hand compatibility evaluation unit 109: Hand determination unit
110: Component gripping position generator, 111 ... Path generator,
112: Hand switching frequency counting unit, 113: Assembly time calculating unit 117: Hand / tool information DB.

Claims

An assembly teaching device for selecting a hand that holds a part in an assembly operation of an assembly,
A storage unit for storing a processing program for selecting a hand and a related database;
A data processing unit for executing the processing program;
With
The related database includes a parts information DB storing information about parts,
A hand information DB storing information on one or more hands;
Including
The data processing unit executes the processing program,
Based on the information in the parts information DB and the hand information DB,
Generating teaching information for selecting the hand holding the part;
An assembly teaching device.
The component information DB includes information indicating the size, mass, position of the center of gravity, material, and surface state of the component,
The assembly teaching apparatus according to claim 1, wherein the hand information DB includes information indicating a hand type, an opening / closing operation range, and a load capacity.
The data processing unit executes the processing program,
Further, a procedure for assembling the assembly and an assembly direction are generated.
Information on the generated assembly direction; and
Based on the information of the component information DB and the hand information DB,
The assembly teaching apparatus according to claim 1, wherein teaching information for determining a gripping position of the part is generated.
In the generation of the teaching information, the data processing unit refers to the information in the component information DB and the hand information DB,
4. Teaching information for selecting the hand is generated based on suitability for at least one physical attribute of mass, gravity center position, material, and surface state of the part and the hand. The assembly teaching apparatus according to any one of the above items.
The assembly teaching device comprises:
In addition, there is a measuring unit equipped with a timer for measuring the number of times the hand to be gripped with respect to the part is replaced or a time required for replacement, or a timer for calculating the sum of the hand replacement time and the assembly time of the part. And
The data processing unit executes the processing program,
5. The hand exchange schedule of the type of hand to be exchanged, the number of hand exchanges, and the hand exchange timing is calculated based on the information measured by the measuring unit. The assembly teaching apparatus according to the section.
An assembly teaching method for selecting a hand that holds a part in an assembly operation of an assembly using a computer,
The computer
By executing a processing program that selects a hand in the data processing unit,
An assembly sequence generation process for generating a procedure for assembling an assembly and an assembly direction;
Based on the parts information DB and hand information DB stored in the storage unit,
A hand selection information generating step for generating teaching information for selecting a hand that holds the component from a plurality of hands;
An assembly teaching method comprising: