JPH11104980A - Robot work route forming-copying method and system device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To absorb a solid difference by using information about the shape of a previously taught route and the shape of an object surface in the periphery of the route relating to a work route divided in a section, and by a search based on a three-dimensional shape data supplied by a sensor. SOLUTION: Corresponding hysteresis information consisting of mapping information or the like by position/attitude conversion of a typical object and individual work object is transfer fed to a work finish section information accumulation part 6. Based on shape detection information and teach information related to work route shape of typical object presented by a route teach information accumulation part 2, based on corresponding hysteresis information presented by the work finish section information accumulation part 6 and solid difference information presented by the route teach information accumulation part 2, in a section during work, correspondence of the typical object and individual object is determined by shape matching technique or the like. In this why, while simultaneously removing an influence of the solid difference and an influence of error or the like in a shape measuring part 3, a work route on a work object is calculated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is used for controlling the following of a work path by a robot having a work tool at a tip thereof, such as welding or deburring, and in particular, based on information on the work path given by a sensor. The present invention relates to a robot work path generation copying method for generating a path to be followed by a work tool, and a system device directly used for implementing the method.

[0002]

2. Description of the Related Art In order to achieve high accuracy in a robot-based route following system, i) information on a work route must be obtained in advance. ii) The generated work route information is accurate. iii) The generated work route is continuous and smooth. Is desired.

[0003] As a conventional technique, a method of generating a path represented by a function based on pre-read sensor data and its reliability using a laser range finder or the like capable of directly measuring a three-dimensional shape ( Japanese Patent Application No. 4-1
24995) has been proposed. However, although the data output from the laser range finder is often superimposed with noise, this method is greatly affected by a shift due to noise or data loss. To address this point, a method of preliminarily teaching the route and the surrounding shape of the representative object, and being tolerant of noise and loss of sensor data when working on each object (Japanese Patent Application No. Hei.
No. 59122) has been invented.

[0004]

In the conventional method, when comparing the position and orientation between each work object and the representative object, it is assumed that the individual difference between the two is small. There are drawbacks to use for target objects with differences.

The main objects to be solved by the present invention are as follows. That is, the first object of the present invention has been made to solve these drawbacks, and can absorb individual differences even when an object having a small individual difference with respect to the representative object is targeted. An object of the present invention is to provide a robot work path generation copying method and a system device.

A second object of the present invention is to provide a robot work path generation copying method and system which suppresses errors in sensor data and generates path information which facilitates work over the entire section of a path on a work object which changes in various ways. No device is provided.

A third object of the present invention is to provide a method for matching and superimposing the teaching information and the detection information, by taking into account the generated and executed copying history information and the prediction information obtained in advance from the teaching-input individual difference information. And a robot work path generation copying method and system apparatus for generating a work path on each work target object while simultaneously removing the effects of the detection information error and the like.

Another object of the present invention is to provide a specification, drawings,
In particular, it will be obvious from the description of each claim in the claims.

[0009]

In order to solve the above-mentioned problems, a route generation copying system device according to the present invention has a route information teaching storage means for storing and outputting teaching information including individual difference information, and a shape for outputting detection information. Measuring means, weighted correspondence determining means for determining and outputting weight information of each work section, work section information accumulating means for accumulating and outputting work route history information in consideration of the weight information, and the extracted individual difference information A path generation unit that fetches the teaching information, the detection information, and the history information, sequentially generates and outputs work path information of a current work section based on the calculated prediction information; Path following means for performing sequential tracking control.

According to the route generation copying method of the present invention, a path to be worked on a representative object among work objects is divided into appropriate sections, and a work path or an object near a work path is divided into each section. It is taught in advance as teaching information on the shape of the object. Here, for the purpose of absorbing individual differences, sections where shapes are likely to appear around a path such as a part with a large curvature and its surroundings are finely divided, and if limited to only these sections, the representative object and the work object The difference is made small, and what kind of individual difference the shape including the surroundings in each section may cause with respect to the work route is taught as individual difference information.

When working on individual objects, not only the route data output from the sensors and the detection information on the shape of the target object near the route, but also these pieces of history information at the part where the work has already been completed are used. Then, the range of the individual difference information is predicted in advance as prediction information, and while using the teaching information on the shape of the work route taught in advance or the shape of an object near the route, the range within the deviation due to the individual difference information is used. The work route of both the detection information and the teaching information and the shape of the periphery of the route are searched for and determined by the least square method or the like.

At this time, it is characterized in that the influence of the shape in each section on the associated parameter is appropriately used. This effect can also be exerted by means such as teaching. After the correspondence is determined in this way, a work route to be worked on is determined one after another, and an accurate work route is adaptively generated for an object to be worked.

[0013] More specifically, in order to solve the above-mentioned problems, the present invention adopts a novel characteristic configuration method and means ranging from a superordinate concept to a subordinate concept listed below. The above objective is achieved.

That is, a first feature of the method of the present invention resides in that a work route on a representative object in a group of work objects and a shape of a representative object near the work route are obtained and prepared in advance, and individual difference information is prepared. The tracking control is performed while sequentially determining and generating a work route by a process of associating teaching information including a work route on each work target object with sensor data and detection information on the shape of each work target object near the work route based on the sensor data. At the time of execution, a work determined from the work history path that has been generated and executed, the history information of the position / posture based on the shape information near the work history path, and the individual difference information extracted from the teaching information. The influence of the individual difference and the sensor data error and the like is simultaneously taken into account, taking into account the position information of the work route to be attempted and the prediction information of the position and orientation of the surrounding shape. While removed by, it is the the configuration adopted for the work target object on the working path robot working path generation copying method comprising calculated.

A second feature of the method according to the present invention is that the working route on the representative object in the first feature of the method according to the present invention is divided into a plurality of sections to adopt a robot working route generation copying method. It is in.

According to a third feature of the method of the present invention, there is provided a robot working path generation copying method according to the second feature of the present invention, wherein the plurality of sections are divided so as to absorb individual differences as much as possible. In the configuration adoption.

A fourth feature of the method of the present invention is that, in the second or third feature of the method of the present invention, a plurality of sections are divided into a work path and a section in which a shape feature is likely to appear around the work path. The present invention resides in adopting a configuration of a robot work path generation copying method.

A fifth feature of the method of the present invention is that the shape features to be distinguished in the fourth feature of the method of the present invention are a straight section and a curved section, a point at which the curvature largely changes, a break point on the work route, and a surrounding area. An object of the present invention is to adopt a configuration of a robot work path generation copying method represented by a point where a qualitative property of a shape changes.

A sixth feature of the method of the present invention is that the work route on the representative object in the second, third, fourth or fifth feature of the method of the present invention is such that the coordinate system appropriately set on the representative object. The present invention employs a configuration of a robot work path generation copying method represented by a curve parameter or the like approximated by a polynomial for each section.

A seventh feature of the method of the present invention is that the teaching information of the representative object shape near the work route in the first, second, third, fourth, fifth or sixth feature of the method of the present invention is: An object of the present invention is to adopt a configuration of a robot work path generation copying method including qualitative information and quantitative information of a shape around each section obtained by dividing the work path.

An eighth feature of the method of the present invention resides in that the qualitative information in the seventh feature of the method of the present invention is such that the qualitative information is represented by a chamfer along a plane, a curve or a work route. In the configuration adoption.

According to a ninth feature of the method of the present invention, the quantitative information in the seventh or eighth feature of the method of the present invention is characterized in that a robot work path generation represented by a parameter of a plane or a numerical expression of each section. It lies in adopting the configuration of the copying method.

According to a tenth feature of the method of the present invention, the teaching information in the first, second, third, fourth, fifth, sixth, seventh, eighth or ninth feature of the method of the present invention is provided. , CA of the representative object
An object of the present invention is to divide a work path on a shape represented by a D model or the like into appropriate sections, and employ a configuration of a robot work path generation copying method which is work path position shape information and information on a peripheral shape.

An eleventh feature of the method of the present invention is the first, second, third, fourth, fifth, sixth, seventh, eighth, ninth, or tenth feature of the method of the present invention. The present invention resides in adopting a configuration of a robot work route generation copying method in which teaching information includes a direction and a range of an individual difference with respect to a work route for each of the divided sections.

A twelfth feature of the method of the present invention is that the first, second, third, fourth, fifth, sixth, seventh, eighth, ninth, tenth, or eleventh of the method of the present invention are described. The process of associating the teaching information with the detection information in the feature of the above is to obtain a position / posture conversion parameter between the representative object and the work target object based on the teaching information and the detection information, and calculate the work route information of the representative object. The present invention resides in adopting a configuration of a robot work route generation copying method applied to the work target object.

According to a thirteenth feature of the method of the present invention, the second, third, fourth, fifth, sixth, seventh, eighth, ninth, tenth, eleventh, or twelfth of the method of the present invention are described. In the conversion process in which the prediction information in the feature of the above maps the position / posture of the teaching information to the position / posture of the detection information, the work history path of the section where the work has been executed and the position / posture of the shape near the work history path An object of the present invention is to employ a configuration of a robot work path generation copying method which is information for predicting a work path to be worked on and a position / posture of a shape near the work path determined based on the history information and the individual difference information.

According to a fourteenth feature of the method of the present invention, the conversion process to be mapped in the thirteenth feature of the method of the present invention is such that the work path for the appropriate range and the look-ahead section in the history part where the work has been completed is performed. Another object of the present invention is to adopt a configuration of a robot work path generation copying method that sequentially obtains conversion so that history information and prediction information of the work path peripheral shape most closely match teaching information.

A fifteenth feature of the method of the present invention is that the first, second, third, fourth, fifth, sixth, seventh, eighth, ninth, tenth, and eleventh aspects of the method of the present invention are described. Twelfth, thirteenth or fourteenth
The history information and the prediction information in the feature (1) are in the configuration adoption of the robot work route generation copying method which is the mapping information.

According to a sixteenth feature of the method of the present invention, the history information in the fifteenth feature of the method of the present invention is characterized in that the history information in each section is determined by the individual difference of the work object and the correlation between the detection information and the teaching information. Another object of the present invention is to adopt a configuration of a robot work path generation copying method in which a weighting process is performed for each section in consideration of different influences on the position / posture determination of an object.

A seventeenth feature of the method of the present invention is that the weighting process in the sixteenth feature of the method of the present invention increases as the weight becomes closer to the section currently being worked on, and the size of the portion of the detection information corresponding to the teaching information. The present invention resides in the adoption of a configuration of a robot work path generation copying method which is enlarged in accordance with the above.

An eighteenth feature of the method of the present invention is that the first, second, third, fourth, fifth, sixth, seventh, eighth, ninth, tenth, and eleventh of the method of the present invention are described. , Twelfth, thirteenth, fourteenth,
When the associating process of the teaching information and the detection information based on the history information and the individual difference information in the fifteenth, sixteenth, or seventeenth feature sequentially superimposes the detection information and the teaching information,
Within the range of deviation due to individual differences from the history information and the individual difference information, a correspondence between a work route of both the detection information and the teaching information and a shape around the work route is determined and determined by a least square method or the like. The present invention resides in adopting a configuration of a robot work path generation copying method.

A nineteenth feature of the method of the present invention is that the first, second, third, fourth, fifth, sixth, seventh, eighth, ninth, tenth, and eleventh of the method of the present invention are described. , Twelfth, thirteenth, fourteenth,
The sensor data in the fifteenth, sixteenth, seventeenth, or eighteenth feature lies in the adoption of a configuration of a robot work path generation copying method that is three-dimensional shape data near a work path by a laser range finder.

A twentieth feature of the method of the present invention resides in that the calculation of the work route on each work object in the first or eighteenth feature of the method of the present invention is performed in a case where a mismatch occurs due to individual differences. Based on the information and the individual difference information extracted from the teaching information, the range of the individual difference in the section is roughly determined, and then the range of the individual difference is searched. The present invention resides in adopting a configuration of a robot work path generation copying method that determines the rotation / translation conversion that best matches and determines the work path.

According to a twenty-first feature of the method of the present invention, the shape coincidence of both the teaching information and the detection information in the twentieth feature of the present invention is based on the difference between the two based on the surface shape approximated by a plane. In this case, the robot working path generation copying method represented by the volume comparison is adopted.

A twenty-second feature of the method of the present invention is that the first, thirteenth, fourteenth, fifteenth, sixteenth and first aspects of the method of the present invention are described.
In the case where the individual difference information in the seventh, eighteenth, nineteenth, or twentieth feature is that the work route is straight and the peripheral shape is flat, the work route direction remains as a degree of freedom, but the posture around the work route is not reliable. If the work path is a curve with a large radius of curvature and it is difficult to teach the surrounding shape, the degree of freedom in the work path direction is determined, but the reliability of the information on the posture is not good, If it is possible that
The curved section is apt to occur in the length method along the work path, and the straight section is hard to occur. The robot work path generation copying method which numerically represents the characteristic tendency represented above is employed.

The first feature of the apparatus of the present invention resides in that path copying means, shape measuring means mounted on the path copying means for measuring a three-dimensional shape of a work object, The work route on the representative object is divided into a plurality of sections, and information on the work route, shapes of objects near the route, and information on individual differences between the representative object and individual objects are taught. Path information teaching storage means for storing, path generation means for generating work path information for the robot, and correspondence history information between the representative object and each object used for path generation in the path generation means. In the robot system having the accumulated work section information accumulating means for accumulating the information, the path generating means may include a path and a path obtained by the shape measuring means in each of the divided sections. Input processing of detection information regarding the peripheral shape, and transfer the association history information including mapping information and the like obtained by converting the position and orientation of the representative object and the individual work target object obtained as a result to the processed section information storage means. Then, based on the shape detection information and the teaching information on the work path shape of the representative object provided by the route teaching information storage means, the association history information provided by the worked section information storage means and the teaching history A robot work path generation copying system configured to be able to freely determine the correspondence between the representative object and each individual object in the section under work based on the individual difference information provided by the information storage means using a shape matching technique or the like. The configuration of the device is adopted.

A second feature of the apparatus of the present invention resides in that the worked section information accumulating means in the first feature of the apparatus of the present invention is adapted to calculate the position / posture information of the object to be worked in each section by the path generating means. In the process, weighted association determining means for weighting mapping information of each section which has been converted and associated, and weighting the position / posture information provided by the route teaching storage means for each corresponding section based on the properties thereof. The present invention is to adopt a configuration of a robot work path generation copying system device in which the weighted position correspondence mapping information is connected to be freely input.

A third feature of the device of the present invention is that the route information teaching storage means in the first or second feature of the device of the present invention is a robot work route constituted by a computer having a graphics function such as a workstation. This is in the configuration and adoption of the generation copying system device.

A fourth feature of the device of the present invention is that the shape measuring means in the first, second or third feature of the device of the present invention emits a planar laser beam and emits a linear laser beam hitting an object to be worked. Another object of the present invention is to adopt a configuration of a robot work path generation copying system device that is configured by a laser range finder that captures a light image.

A fifth feature of the apparatus of the present invention is that the path copying means in the first, second, third or fourth feature of the above-mentioned apparatus of the present invention is a robot work path generating copying system apparatus comprising an industrial robot. In the configuration adoption.

A sixth feature of the device of the present invention is that the route information teaching storage means in the third, fourth or fifth feature of the device of the present invention operates the shape measuring means for a representative object. A robot work path generation copy constructed by displaying detected data or teaching CAD data relating to a representative object and performing various operations on the graphics screen for the displayed shape. The configuration of the system unit is adopted.

A seventh feature of the device of the present invention resides in that the route information teaching and accumulating means in the first, second, third, fourth, fifth or sixth feature of the device of the present invention is adapted to include a prototype object or a representative object. The shape of the vicinity of the work route is measured by a shape measurement unit as appropriate for the object, and the data obtained by the route information teaching storage unit is displayed by a plurality of teaching units in an expression format that is easy for the instructor to understand. Another object of the present invention is to adopt a configuration of a robot work route generation copying system device to which an information input means capable of inputting a route shape, an object shape around the work route, and information related thereto is added.

An eighth feature of the apparatus of the present invention is that the shape matching technique in the first feature of the apparatus of the present invention is such that mapping transformation, coordinate transformation, transformation parameter calculation, least square method, rotation / translation transformation, The present invention resides in the adoption of a configuration of a robot work path generation copying system device represented by volume calculation.

A ninth feature of the device of the present invention resides in that the route generating unit in the first, second, third, fourth, fifth, sixth, seventh or eighth feature of the device of the present invention has already been operated. A section information storage unit is configured to employ a robot work route generation copying system apparatus integrally configured by a computer.

[0045]

Embodiments of the present invention will be described with reference to the drawings. FIG. 1 is an external perspective view of an example of a work target object and a work path example, FIG. 2 is a diagram showing an example of a work system path on a representative object, and FIGS. FIG. 4 is a diagram illustrating a state explanatory diagram and an example of detection information thereof, FIG. 4 is a diagram illustrating a procedure for generating a route based on data output from an external sensor, and FIG. 5 is a diagram illustrating a procedure for generating a work route based on data output from an external sensor. It is a figure which produces | generates a work path | route also using the sensor data of the work completed area of an appropriate range.

FIG. 1 shows an object A to be worked and a work route γ.
Here is an example. From the three-dimensional shape data near the route α obtained from the sensor for such a work route γ, the work route γ
Is determined based on the shape detection information obtained from the sensor data, a conversion parameter of the position / posture between the work target object A and the representative object B is obtained, and the path β of the target object B is determined.
May be applied to the work object A.

As shown in FIG. 1, the shape of the path γ to be worked may be variously changed. Therefore, first, as shown in FIG.
The work route β expressed by the CAD model or the like is divided into appropriate sections, and the route position shape β ′ information and the teaching information on the peripheral shape β ″ are taught and stored.

Here, the work route β is divided into, for example, a straight section and a curved section, a point at which the curvature greatly changes, and a path β.
It is divided at the upper break point or at a point where the qualitative property of the surrounding shape changes, and each section is represented by a curve approximated by a polynomial and stored.

For the peripheral shape, qualitative information (plane, curved surface, etc.) and quantitative information (plane parameters, etc.) of the peripheral shape are taught and stored for each section. In addition, the instructor also determines what individual difference the route β has with respect to the route β, and inputs the direction and range of the individual difference information.

The operations such as welding and deburring are performed while sequentially proceeding along the work path γ on the work object A. During execution, data on the position of the path α from the sensor 1 and data on the shape of the object near the path α Is output. A sensor often used for such an application is a laser range finder, and output data thereof is as shown in FIG.

In the example shown in FIG. 3A, a flat laser beam L is emitted from the sensor 1 and an image of the linear light L hitting the work object A is captured by the sensor 1, so that the linear portion of the linear portion is captured. A three-dimensional shape is calculated. By recognizing a characteristic shape of the path section as shown in the data example of FIG. 3B, the path position αn in the section is calculated.

However, since data often includes missing portions and noise, it may be difficult to calculate the route position αn accurately from the data as shown in FIG. In order to deal with such a case, the teaching information of the shape taught in advance for the route shape α ′ and the shape α ″ around the route is used. Considering this, as shown in FIG. 5, a work route γ is generated for a section in which work has already been performed, and data on the shape near the route γ has been obtained.

Accordingly, the position and orientation of the work path β and the data of the surrounding shapes β ′ and β ″ previously taught by the teaching information and the history information shown in FIG. It is possible to obtain a transformation that maps to α and the position / posture of the surrounding shapes α ′ and α ″. By this conversion, the robot can know in advance the position information of the work route ε to be worked and the information of the surrounding shape. When the prediction data of the section to be worked on is input from the sensor 1, the position information of the work path γ obtained earlier and the information on the surrounding shape are compared with each other to include the error in the sensor 1 data described above. The path γ can be generated by removing the portion and compensating for the data missing portion.

However, in general, there is often an individual difference between the representative object B and the work object A, and there are many cases where a displacement or deformation occurs during the work. For this reason, even if the position and orientation of the representative object B are converted by the above-described mapping, it may not completely match the work target object A. In the present embodiment, in order to cope with such a case, as shown in FIG. 5, an appropriate range of the work route γ portion where the work has already been completed and a route γ,
The conversion is sequentially determined so that the ε data and the shape data near each of the paths γ and ε best match the teaching information.

Here, when determining the correspondence between the shape detection data of the currently working section and the shape information of the representative object B,
It is necessary to perform a process of determining the correspondence that best matches the range of the individual difference information generated in the previous section. In other words, it is necessary to search for the degree of freedom corresponding to the individual difference information of the previous section, which is greatly affected by the shape of each section and can be predicted. Therefore, by adding such individual difference information when teaching the shape information by dividing the representative object B into sections, even when individual differences occur, the path γ is appropriately removed while removing the influence thereof. Can decide.

[0056]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described with reference to the drawings with reference to an example of an apparatus and an example of a method thereof. In this embodiment, an example of an apparatus and an example of a method will be described as a representative, but this embodiment is not necessarily limited to only the means of the example of the apparatus and the method of the example of the method. The present invention achieves the object of the present invention and can be carried out with appropriate modifications within a range having the effects described below.

(Example of Apparatus) FIG. 6 is a block diagram showing an example of the present apparatus. This device example includes a route information teaching storage unit 2, a shape measurement unit 3, a route generation unit 4, a weighted association determination unit 5, a completed section information storage unit 6, and a route copying unit 7.

The route information teaching storage unit 2 divides the work route β on the representative object B in the group of work objects A into a plurality of sections, and obtains information on the work route β and the shape of the object B near the route. has a function of teaching and accumulating information on individual differences between the representative object B and the individual objects A. The shape measuring unit 3 is mounted on the path copying unit 7 and It has a function to measure a three-dimensional shape.

In the process of calculating the position / posture information of the work object A in each section in the path generation section 4, the weighted association determination section 5 inputs the mapping information converted and associated, and It has a function of weighting the position / posture teaching information provided from the route teaching storage unit 2 based on the nature thereof. The work completed section information storage unit 6 receives the weighted position correspondence mapping information from the weighted correspondence determination unit 5, and receives the representative object B used for the generation of the work path γ in the path generation unit 4 and each of the representative objects B. It has a function of accumulating association history information with the object A.

The route generation unit 4 has a function of generating work route γ information for the robot, and a work route α and a route α obtained by the shape measurement unit 3 in each of the divided sections.
Input processing of detection information about the peripheral shapes α ′, α ″,
An individual work object A with the resulting representative object B
Is transferred to the worked section information storage unit 6, and the shape detection information and the work route β of the representative object B provided from the route teaching information storage unit 2 are transferred. Based on the association history information provided by the worked section information storage unit 6 based on the teaching information on the shape and the individual difference information provided by the route teaching information storage unit 2, the representative object B for the working section is displayed. Is determined by using a shape matching technique or the like, thereby calculating the work path γ on the work object A while simultaneously removing the influence of the individual difference and the influence of the error of the shape measurement unit 3. .

The route information teaching storage unit 2 can be constituted by a computer having a graphics function such as a workstation, and the route generation unit 4 and the worked section information storage unit 6 can be realized by a computer. Also, the shape measuring unit 3
Uses a laser range finder, and the path follower 7 can be constituted by an industrial robot.

(Example of Method) An example of the method of the present embodiment applied to the above example of the apparatus will be described in detail with reference to the drawings. FIG. 7 is a diagram illustrating an example of a section in which the work route and the shape of the periphery of the route have different effects on the route determination parameter.

(1) Step of teaching procedure based on representative object B First, using the path information teaching storage unit 2, the work path β and the shape of the surface of the object B in the vicinity of the representative object B of the work objects A Here, the path information teaching storage unit 2 displays data obtained by operating the shape measuring unit 3 on the representative object B as detection information, or relates to the representative object B. The teaching can be performed by displaying CAD data and performing various operations on the displayed shape on the graphics screen.

Therefore, the work path β is divided into a section where the shape β ′ and the surrounding shape β ″ change greatly or a section to the prayer point on the path β as one section, and the entire work path β is divided into a plurality of sections. Then, an appropriate coordinate system is set on the representative object B, and each section can be expressed on the coordinate system by a parameter of a curve approximated by a polynomial for each section.

The teaching information on the surface of the object B in the vicinity of the work route β can be taught by qualitative information such as a flat portion and a curved surface portion, and chamfering along the route β. After a coordinate system is set, the parameters of these planes and numerical values as in Japanese Patent Application No. 5-114998 can be used.

After teaching the work path β and the surrounding shapes β ′, β ″ of the representative object B in this way, for each section, it is determined how the sensor data of the section affects the position / posture determination of the entire object B. For example, the path β direction remains as the degree of freedom when the path β is a straight line and the surrounding shape is a set of planes and individual differences are taken into consideration, as in the part βa in FIG. 7, but the attitude around the path β is determined with high reliability. Is done.

On the other hand, when the path β is a curve having a large curvature and it is difficult to teach the surrounding shape β ″ as in the part βb, the degree of freedom in the direction of the path β is determined, but the reliability of the individual difference information regarding the posture is good. Also, it teaches what kind of individual difference may occur in the section.For example, a section such as the portion βa has an individual difference in the length direction along the path β. Although it is easy, the section such as the portion βb is unlikely to cause an individual difference.

(2) Work execution procedure stage for each work target object A The shape measuring unit 3 measures a three-dimensional shape near the route α. This can be constituted by a sensor 1 typified by a laser range finder, for example, and the output of the sensor 1 can obtain the position of the work path α and the surrounding shapes α ′, α ″. It is sent to the generation unit 4.

When generating the route γ for the currently working section, the route generating section 4 measures the shape based on the association information of the section where the work has been completed, which is supplied from the worked section information storage section 6. The detection information of the shape data transferred from the section 3 is superimposed on the teaching information on the shape of the currently working section supplied by the path information teaching storage section 2.

Here, when a discrepancy due to the individual difference occurs, the individual difference information taught in advance for the section before the section is extracted, and the range of the individual difference in the section is roughly determined. The range of the difference is searched for, the rotation / translation conversion that makes the two shapes best match is determined, and the path γ is determined. Here, as a measure of the coincidence of the shapes, for example, a method using the volume of the difference between the two based on the surface shape approximated to a plane may be considered.

When the path γ generation for the section currently being worked on is completed in this way, it is stored in the worked section information storage unit 6 together with the history sensor data of the portion corresponding to the history information of the position / posture conversion. The path copying unit 7 performs a copying operation based on the work path γ output by the path generation unit 4.

Also, due to the existence of individual differences in the object A and the relationship between the sensor 1 data obtained by the shape measuring section 3 and the teaching information supplied from the path information teaching storage section 2, each section is defined by the object A. Considering that the influence on the position / posture determination is different, first, for example, the route information teaching storage unit 2
In the above, the instructor judges these effects and teaches the numerical value of the weight, and the weighted value transferred from the route information teaching storage unit 2 in the weighted association determination unit 5 and the section output from the route generation unit 4 The rotation / translation conversion is calculated by a calculation method in which a weighting process is added to the association information of, and is transferred to the worked section information storage unit 6.

This weighting is increased, for example, as it approaches the section currently being worked on. Also, of the sensor 1 data obtained by the shape measuring section 3, the weight of the portion corresponding to the teaching information supplied from the path information teaching storing section 2 is obtained. What is necessary is just to enlarge according to a magnitude | size. In the route information teaching storage unit 2 in FIG.
Range data and object B measured by shape measuring unit 3
It is also possible to additionally provide a means for interactively inputting the work paths α and β and the peripheral shapes α ′, α ″, β ′ and β ″ on the screen. .

[0074]

As described above, according to the present invention, since the actual work object often has individual differences from the representative object, the work path divided into sections has the same shape as the previously taught path. Using information on the shape of the object surface around the route,
By performing a search based on the three-dimensional shape data supplied by the sensor, it is possible to generate a work route that can improve the work quality even in such a case.

According to the present invention, the work can be performed even when the sensor data includes a missing part, so that the number of teaching points can be greatly reduced. Can be adapted to the break points. In addition, the weighting process further relates to the shape.
The present invention has excellent effects, such as greatly reducing the labor required for teaching the user.

[Brief description of the drawings]

FIG. 1 is an external perspective view of a work object example and a work path example.

FIG. 2 is a diagram illustrating an example of a work path of a representative object.

FIGS. 3A and 3B are diagrams illustrating an input state of a work target object by detection and an example of detection information thereof. FIG.

FIG. 4 is an explanatory diagram of a procedure for generating a route based on data output by an external sensor.

FIG. 5 is a diagram illustrating a process of generating a work route using sensor data of a work-completed section in an appropriate range in accordance with a procedure of generating a route based on data output from an external sensor.

FIG. 6 is a block diagram showing an example of an apparatus according to an example of the embodiment of the present invention.

FIG. 7 is a diagram showing an example of a section in which the influence of the work route and the shape of the periphery of the route on the parameters for determining the route is different as an example of the above method.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Sensor 2 ... Path information teaching storage part 3 ... Shape measurement part 4 ... Path generation part 5 ... Weighted association determination part 6 ... Worked section information storage part 7 ... Path copying part A ... Work target object B ... Representative object .alpha., .beta., .gamma., .epsilon. work path .alpha. ', .beta.'... path shape .alpha. ", .beta."... Path β with large curvature

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Yukihiro Nakamura 3-19-2 Nishi-Shinjuku, Shinjuku-ku, Tokyo Japan Telegraph and Telephone Corporation (72) Inventor Makoto Mizukawa 3- 19-2, Nishi-Shinjuku, Shinjuku-ku, Tokyo No. Japan Telegraph and Telephone Corporation

Claims (31)

[Claims] 1. A sensor device comprising: teaching information prepared and obtained in advance on a work path on a representative object of a group of work target objects and a shape of a representative object near the work path and including individual difference information; In performing the profiling control while sequentially determining and generating the work path by associating the work path on each work object with the detection information regarding the shape of each work object in the vicinity of the work path, the generated profiling has been executed. Work history path and position / posture history information based on shape information near the work history path;
The individual difference information extracted from the teaching information, and taking into account the position information of the work route to be worked on and the predicted information of the position and orientation of the surrounding shape calculated based on the individual difference information, the individual difference and the sensor data error etc. A method of calculating a work path on each of the work target objects while simultaneously removing the influence. 2. The method according to claim 1, wherein the work path on the representative object is divided into a plurality of sections. 3. The method according to claim 2, wherein the plurality of sections are divided so as to absorb individual differences as much as possible. 4. The method according to claim 2, wherein the plurality of sections are divided into a work path and a section in which a shape feature is likely to appear around the work path. 5. The shape features to be classified are represented by straight sections and curved sections, points where the curvature changes significantly, break points on the work route, and points where the qualitative properties of the surrounding shape change. The robot work route generation copying method according to claim 4. 6. The work path on the representative object is represented by a curve parameter or the like which is approximated by a polynomial for each section on a coordinate system appropriately set on the representative object. 6. The robot work route generation copying method according to any one of claims 3, 4, and 5. 7. The teaching information of a representative object shape in the vicinity of a work route includes qualitative information and quantitative information of a shape around each section obtained by dividing the work route. 7. The robot work path generation copying method according to 1, 2, 3, 4, 5, or 6. 8. The method according to claim 7, wherein the qualitative information is represented by a chamfer along a plane, a curve, or a work path. 9. The robot work path generation copying method according to claim 7, wherein the quantitative information is represented by a parameter of a plane or a numerical expression of each section. 10. The teaching information is characterized in that a work path on a shape represented by a CAD model or the like of a representative object is divided into appropriate sections, and work path position shape information and information on peripheral shapes are provided. Claims 1, 2, 3, 4, 5, 6, 7,
10. The robot work path generation copying method according to 8 or 9. 11. The teaching information according to claim 1, wherein the teaching information includes a direction and a range of an individual difference with respect to a work route for each of the divided sections.
11. The robot work route generation copying method according to 8, 9, or 10. 12. A process for associating teaching information with detection information, comprising: obtaining a position / posture conversion parameter between a representative object and a work target object based on the teaching information and the detection information; The information is applied to the work object, wherein the information is applied to the work target object.
12. The method according to claim 8, 9, 10 or 11, wherein 13. The prediction information includes: a conversion process for mapping the position / posture of the teaching information to the position / posture of the detection information; The prediction information of a work route to be worked on and a position / posture of a shape near the work route determined based on the history information of the posture and the individual difference information. 5, 6, 7, 8,
13. The robot work path generation copying method according to 9, 10, 11 or 12. 14. The transformation process for mapping is such that the history information and the prediction information of the work route and the shape around the work route in the appropriate range and the pre-reading section in the history part where the work is completed most closely match the teaching information. 14. The method according to claim 13, wherein the conversion is performed sequentially. 15. The system according to claim 1, wherein the history information and the prediction information are mapping information.
The robot working route generation copying method according to 8, 9, 10, 11, 12, 13, or 14. 16. The history information is calculated for each section in consideration of the individual differences of the work object and the different effects on the position / posture determination of the work object in each section due to the correlation between the detection information and the teaching information. The robot work path generation copying method according to claim 15, wherein a weighting process is performed on the robot work path. 17. The weighting process according to claim 16, wherein the weighting process is increased in proportion to a section currently being worked on, and is increased in accordance with the size of a portion of the detection information corresponding to the teaching information. Robot work path generation copying method. 18. A process for associating teaching information with detection information based on history information and individual difference information, comprising the steps of: when superimposing the detection information and the teaching information successively, on the basis of the history information and the individual difference information; 3. The method according to claim 1, wherein a correspondence between a work route of both the detection information and the teaching information and a shape around the work route is determined by a least square method or the like within a range of deviation due to individual differences. 3, 4, 5, 6, 7,
8. The robot working route generation copying method according to 8, 9, 10, 11, 12, 13, 14, 15, 16 or 17. 19. The sensor data according to claim 1, wherein the sensor data is three-dimensional shape data near a work route by a laser range finder.
8, 9, 10, 11, 12, 13, 14, 15, 16,
19. The method of copying a robot work route according to claim 17 or 18. 20. A method of calculating a work route on each work target object, comprising the steps of: when a mismatch occurs due to an individual difference, determining an individual in the section based on history information and individual difference information extracted from teaching information; After the range of the difference is roughly determined, the range of the individual difference is searched, and the rotation / translation conversion is determined so that the shape of both the teaching information and the detection information is most matched to determine the work path. The method according to claim 1 or 18, wherein 21. The method according to claim 20, wherein the shape matching of both the teaching information and the detection information is represented by a volume comparison of both difference portions based on a surface shape whose scale is approximated by a plane. Robot work path generation copying method. 22. The individual difference information indicates that, when the work path is straight and the peripheral shape is flat, the work path direction remains as a degree of freedom, but the posture around the work path has high reliability, and the work path has a radius of curvature. When teaching the surrounding shape with a large curve is difficult, the degree of freedom in the direction of the work path is determined, but the reliability of the information on the posture is not good, and there is a possibility of what kind of individual difference occurs in the section. The curve section is likely to occur in the length method along the work route and the straight section is unlikely to occur, and the characteristic tendency represented above is numerically expressed. 16,
21. The robot working path generation copying method according to any one of 17, 18, 19, and 20. 23. A path following means, a shape measuring means mounted on the path following means for measuring a three-dimensional shape of a work object, and a work on a representative object among the work objects. A path information teaching storage unit that divides the path into a plurality of sections, teaches and stores information on the work path, the shape of an object near the path, and information on individual differences between the representative object and the individual objects. ,
Path generation means for generating work path information for the robot; and worked section information storage means for storing history information relating the representative object used for the path generation to the individual objects in the path generation means. In the robot system having the above, the path generating means, in each of the divided sections, performs input processing of the detection information on the path or the shape around the path obtained by the shape measuring means, and the representative object obtained as a result and the individual The association history information including mapping information and the like obtained by converting the position / posture with the work target object is transferred to the worked section information storage unit, and the shape detection information and the representative object provided by the route teaching information storage unit are transferred. The correspondence history information provided by the worked section information storage means and the teaching information storage means Robot work path generation simulation, wherein the correspondence between the representative object and the individual object in the section under work is freely determined by shape matching technology or the like based on the individual difference information provided by the robot. System device. 24. The work section information storage means, in the process of calculating the position / posture information of the work target object in each section by the route generation means, stores mapping information of each section which has been converted and associated, and The position and mapping information weighted by the weighting determination unit that weights the position and orientation information provided by the route teaching storage unit based on the nature of the section is connected to each of the sections so as to be freely input. 24. The robot work path generation copying system apparatus according to claim 23, wherein: 25. The robot work route generation copying system apparatus according to claim 23, wherein the route information teaching storage means is constituted by a computer having a graphics function such as a workstation. 26. The apparatus according to claim 23, wherein said shape measuring means comprises a laser range finder which emits a planar laser light and captures an image of a linear light hitting an object to be worked. The robot work route generation copying system device according to [1]. 27. The apparatus according to claim 23, 24, 25 or 26, wherein the path copying means is constituted by an industrial robot. 28. A path information teaching storage means for displaying detection data obtained by operating a shape measuring means on a representative object, or teaching CAD data on a representative object, and 28. The robot work path generation copying system apparatus according to claim 25, 26 or 27, wherein teaching is performed by performing various operations on a graphics screen. 29. A route information teaching and accumulating means for measuring a shape near a work route by a shape measuring means as appropriate for a trial object or a representative object, and converting the data obtained by said route information teaching and accumulating means into a plurality of teaching means. Information input means for displaying the work path shape, the object shape around the work path and information relating to the work path shape, the work path shape, the work path shape, and the like. Claims 23, 24, 25, 26, 27
Or the robot work route generation copying system apparatus according to 28. 30. The shape matching technique according to claim 23, wherein the shape matching technique is typified by mapping conversion, coordinate conversion, conversion parameter calculation, least squares method, rotation / translation conversion, and volume calculation of a difference part. Robot work path generation copying system device. 31. The apparatus according to claim 23, wherein the route generating section and the worked section information accumulating section are integrally configured by a computer.
The robot work path generation copying system device according to 7, 28, 29 or 30.