JPH0934528A - Robot control method and robot system device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a robot control method and a robot system device which can move a manipulator to its accurate target path even when an external sensor cannot function when the target path is produced by the external sensor. SOLUTION: When the locus information on a manipulator is produced and the operation of a robot is controlled by a robot controller, a teaching path is generated by a teaching path generation part 70 based on the discrete sequence-of-point information on a taught work object. Then a sensing path is generated by a sensing path generation part 71 based on the discrete sequence- of-point information on the work object that is acquired by a sensing device. A target path generation part 72 maps both teaching and sensing paths together to generate a target path. Then a locus generation part 73 generates the locus information on the manipulator based on the target path.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a robot control method suitable for copy control of a work robot and a robot system apparatus for carrying out the same, and more particularly to a robot control method which enables highly accurate machining by a robot and The present invention relates to a robot system device.

[0002]

2. Description of the Related Art This type of robot control method is called a teaching playback method, and comprises a teaching device 1, a robot control device 2, a servo drive device 3 and a manipulator 4 having a multi-axis structure as shown in FIG. It is realized by the robot system device β.

In the robot control device 2, the teaching device 1 such as a teaching pendant is used to teach a work route or the like, and information on a teaching point sequence is obtained to drive the servo drive device 3 in the following procedure.

First, as shown in FIG. 12, the information of the teaching point sequence is converted into a spatial trajectory Xref represented in discrete time, which is composed of a set of the position, orientation and time of the hand of the manipulator 4 in a "transform" block ( C'1). Then the spatial orbit
Angle information θref of the joint space is generated based on Xref (C′2), and this is sent to the servo drive device 3.

Upon receiving the joint space angle information θref from the robot controller 2, the servo drive unit 3 controls the drive of the servomotor so that the position and orientation of the manipulator 4 reaches the joint angle θref (C'3). ).

Therefore, the conventional robot system device β is configured to control the driving of the manipulator 4 based on the information of the teaching point sequence. An external sensor may be used to confirm the target position of the work object, but such sensor is controlled via the input / output port and functions as an external device.

[0007]

By the way, in the conventional robot system apparatus, even if the work route is taught, the resolution of the internal sensor for detecting the position and orientation of the manipulator is a finite value, and the actual work target Since there are individual differences and installation errors between the object (workpiece) and the teaching object (masterpiece), it is difficult to drive and control the manipulator with high accuracy along the work path.

On the other hand, there is also a robot control method in which the manipulator is driven by following the work path with respect to the work object by an external sensor. However, in such a control method, the measurement range may be limited due to the limitation of the measurement range in the distance measurement sensor or the dependency of the optical property of the measurement target in the optical sensor, so that the continuation of the driving of the manipulator is not guaranteed. is there.

Here, the main objects to be solved by the present invention are as follows. A first object of the present invention is to provide a robot control method and a robot system device capable of driving a manipulator to a target path with high accuracy.

A second object of the present invention is to make a manipulator follow an accurate target path even when the external sensor falls into an unmeasurable state when a target path is created using the external sensor. An object of the present invention is to provide a robot control method and a robot system device which can be performed.

A third object of the present invention is to provide a robot control method and a robot system apparatus capable of finely controlling a manipulator even if coarse teaching is performed in the copying control and reducing the labor of teaching. Is.

[0012] Other objects of the present invention are as follows:
In particular, it will be obvious from the description of the appended claims.

[0013]

[Means for Solving the Problems] To solve the above-mentioned problems,
The above object is achieved by the present invention employing the following novel characteristic configuration methods and means.

That is, the first feature of the method of the present invention is that when the robot control means creates the trajectory information of the manipulator to control the operation of the robot, the sensing information and the teaching information for the work object are simultaneously and in parallel. While inputting and associating the two pieces of information, the sensing information is preferentially adopted, and when the sensing information is interrupted during the work, a mapping process that captures the teaching information at that time is performed, and the work is consistently performed. This is a robot control method in which continuous smooth execution is performed.

The second feature of the method of the present invention is that when the robot control means creates trajectory information of the manipulator and controls the operation of the robot, the manipulator and the servo that drives the manipulator based on the trajectory information input. Discrete points relating to the work object taught by the teaching means, including driving means, teaching means for teaching work by the manipulator, sensing means for detecting external world information, and the robot control means. A teaching route is generated from the row information, and a sensing route is generated from the discrete point sequence information related to the work object obtained by the sensing means, and the sensing route and the teaching route are mapped to obtain a target route. Is generated, and trajectory information of the manipulator is generated from the target path.

A third feature of the method of the present invention is that the discrete point sequence information obtained from the teaching means in the second feature of the method of the present invention represents a space curve by giving a plurality of control points and an interpolation method. It is in the robot control method that becomes numerical information.

A fourth feature of the method of the present invention is that the mapping process in the first, second or third feature of the method of the present invention is such that the sensing success point is the sensing path and the sensing failure point is the target path. The robot control method is a target route that is subjected to geometric transformation processing such as parallel movement, congruential transformation, similarity transformation, and affine transformation so that the teaching route is coupled to the sensing route.

A fifth feature of the method of the present invention is a robot control method in which the manipulator according to the first, second, third or fourth feature of the method of the present invention has a structure having a plurality of axes.

The sixth feature of the method of the present invention is that the generation of the sensing path in the second, third, fourth or fifth feature of the method of the present invention is feature point processing for detecting feature points of a work object. And the detected feature points are used as reference points when creating a route.

The seventh feature of the method of the present invention is that the movement amount of the manipulator is used to create trajectory information of the manipulator in the second, third, fourth, fifth or sixth feature of the method of the present invention. In the detection, there is a robot control method in which a velocity profile is created based on the velocity pattern of the hand of the manipulator and the maximum velocity, and the current movement amount of the manipulator is determined by the velocity profile and the elapsed time.

The first feature of the device of the present invention is a manipulator, a servo drive device for driving the manipulator based on the input trajectory information, a teaching device for teaching work by the manipulator, and a sensing device for detecting external information. In a robot system including a robot control device, the robot control device includes a teaching route generation unit that generates a teaching route from discrete point sequence information related to a work target taught by the teaching device, From the sensing route generation unit that generates a sensing route from the discrete point sequence information of the work target detected by the sensing device, the target route generation unit that maps the sensing route and the teaching route, and the target route And a trajectory generation unit that generates trajectory information of the manipulator.

A second feature of the device of the present invention is that the trajectory generator in the first feature of the device of the present invention creates a velocity profile based on the velocity pattern of the hand of the manipulator and the maximum velocity, and The robot system device is connected to an S generator that generates the amount of movement of the manipulator from the velocity profile and the elapsed time, and is configured to obtain information on the amount of movement of the manipulator at the present time.

A third feature of the device of the present invention is a robot system device in which the external sensor of the sensing device according to the first or second feature of the device of the present invention is arranged with the tip of the manipulator being offset. It is in.

A fourth feature of the device of the present invention resides in a robot system device in which the manipulator according to the first, second or third feature of the device of the present invention has a structure having a plurality of axes.

[0025]

The present invention takes the following actions because it takes the novel method and means as described above.

According to the present invention, since the sensing information and the teaching information are used to create the target route, the robot controller can control the target position / orientation of the hand of the manipulator from moment to moment, and the accuracy of the contour control of the curved shape can be improved. It enables highly accurate robot control. Further, it is possible to coarsely teach and perform fine copying control.

In particular, when the feature point of the work target cannot be detected from the information measured by the external sensor and the sensing fails, the robot is determined from the sensing information given up to that point and the teaching information previously taught. It is possible to determine the position and orientation to be taken by the minions.

[0028]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in more detail below with reference to the accompanying drawings on the basis of an apparatus example and a method example thereof. In the present apparatus example and method example, the “path to be followed” by the robot system apparatus α will be described as a curved shape (see FIG. 4) generated by the curved surface unevenness of the work target A.
The present invention is not limited to these work targets and the like.

(Apparatus Example) FIG. 1 shows a main configuration of a robot system apparatus of the present apparatus example, FIG. 2 shows a mounting position of an external sensor, and FIG. 3 shows a main part of a robot controller. It should be noted that the same reference numerals are used for the parts having the same names as those of the configuration of the robot system device β described in the section of the related art.

As shown in FIG. 1, the robot system device α comprises a teaching device 1, a robot control device 2, a servo drive device 3, a manipulator 4 and a sensing device 5. The teaching device 1 serves to teach the robot controller 2 the work by the manipulator 4 by the teaching playback method.

The sensing device 5 has an external world sensor 5a for obtaining external world information. As shown in FIG. 2, the external sensor 5a is arranged at a position offset from the hand of the manipulator 4.

The servo drive device 3 functions to drive the manipulator 4 based on the input trajectory information. The manipulator 4 has a multi-axis structure and is driven by the servo drive device 3.

The robot controller 2 has the feature point processing unit 6 and the Xref generator 7 shown in FIG. 3 as main components, and additionally has a memory and a control unit (not shown).

The characteristic point processing unit 6 obtains the coordinate information of the actual route and the additional information transmitted from the sensing device 5 at a constant time interval by sensing. Then, the coordinate information of the actual route is converted into the world coordinate system (coordinate system based on the floor surface) to create a sensing information sequence, and the information of the sensing information sequence is sent to the sensing route generation unit 71 of the Xref generator 7. Send out.

As for the feature point detection algorithm of the object in the feature point processing unit 6, the basic invention such as Mr. Kaneko's shape processing method (Japanese Patent Application No. 4-114428) is used. For example, when a one-dimensional range sensor is used as the sensing device 5, as shown in FIG. 4, the distance is measured in the ab direction that intersects the curved shape of the object A and is shown in the graph of FIG. Obtain cross-sectional shape information. Then, the feature point (point c in FIG. 5) and the spatial coordinates of the feature point are determined from the cross-sectional shape information. The boundary (actual path) B of the mating surface is detected by continuously executing such a procedure along the curved shape of the object A.

The Xref generator 7 of the robot controller 2 is
As shown in FIG. 3, the teaching route generation unit 70, the sensing route generation unit 71, the target route generation unit 72, the trajectory generation unit 73,
It is composed of a velocity profile section 74 and an S generator 75.

The teaching path generating unit 70 converts the input teaching point sequence information from the discrete teaching point sequence into the teaching route Tw (s) represented by the length (S) of the curve, and the target route generating unit 72. Send to. The teaching path Tw (s) at this time is a three-dimensional vector.

The information on the teaching point sequence is expanded from the conventional one consisting of information on a start point and an end point and a path (straight line, arc, etc.) connecting them, and a plurality of control points consisting of a discrete teaching point sequence. An interpolation method was given to represent the space curve.
For example, if two control points are given and the interpolation method is a straight line, a straight line connecting the two points is given as in the conventional case, and if a spline is given as a plurality of control points and an interpolation method, a smooth space curve is given. .

The sensing path generation unit 71 converts the points of the discrete sensing information string on the path to be copied into the sensing path Sw (s) by connecting the points according to the shape (interpolation method) of the teaching path given by the teaching. Then, this is sent to the target route generation unit 72.

The target path generation unit 72 associates the sensing path Sw (s) for the actual work object (workpiece) with the teaching path Tw (s) for the teaching object (masterpiece). Here, the teaching path Tw (s) of the master piece and the sensing path Sw (s) of the workpiece
Are assumed to be connected by any relation of translation, geometric transformation such as congruential transformation, similarity transformation, and affine transformation.

That is, it is assumed that they are connected by the following general formula. Sw (s) = A · Tw (s) + B (1) Here, A is a 3 × 3 matrix representing linear transformation (rotation), and B is a three-dimensional vector representing translation. The method of obtaining A and B will be described later.

According to the above correspondence, as shown in FIG. 10, when the sensing device 5 succeeds in sensing, the sensing path Sw (s) is used, and when the sensing fails (sensing miss section), the teaching path Tw (s). If the mapping process is performed using (), the target route Sw ′ (s) can be generated as in the following equation.

[0043]

[Equation 1]

In this example of the apparatus, the working posture of the manipulator 4 is the posture at the time of teaching by the teaching device 1.

The trajectory generation unit 73 has the target route Sw '(s) and the velocity profile unit 7 sent from the target route generation unit 72.
4 and the Xref of the manipulator to be sequentially controlled based on the movement amount of the manipulator 4 created by the S generator 75.
Generate orbital information of. Details of such data conversion will be described later.

The speed profile unit 74 generates a speed profile using the speed pattern of the manipulator 4 (the shape of speed when time is a variable as shown in FIG. 6) and the maximum speed as parameters. In the contact copying as in the example of the present device, a force sensor (not shown) is provided, and a change in speed or the like is monitored by the value of the force sensor to correct the speed profile.

The S generator 75 uses the velocity profile as a function of the movement distance S of the manipulator 4, and creates the value of S at the current time.

Next, a robot operation command used in the robot system apparatus α of this apparatus example will be described. 7 and 8 each show a control model of a robot control method used in the robot system apparatus α of this apparatus example. It should be noted that the hand to be used by omission refers to the tip of the manipulator 4, and the hand path designation refers to designation of a path along which the tip of the manipulator 4 moves.

(Robot operation command) The robot operation commands derived from the control model in this device example are: A) world coordinate system and joint coordinate system B) presence / absence of external sensor 5a C) hand path (teaching) using the teaching device 1 It is determined by the combination of whether or not the route is specified.

However, in this example of the apparatus, the drive control of the manipulator 4 is performed with reference to the world coordinate system. Therefore, in the case of the joint coordinate system, the external sensor 5a is used only when the external sensor 5a is not used and the hand path is not specified. Since "Yes" requires a route to be corrected, only the case where the fingertip route is specified is targeted. When the external sensor 5a is not used, it is equivalent to the conventional robot system device β.

1) Joint coordinate system / no external sensor 5a used / no hand path designation In this case, the teaching point sequence in FIG. 7 or 8 is the target joint angle, and the velocity profile and maximum velocity are used to create trajectory information. Given that the transfer function of the Xref generator 7 is 1 (that is, the input is output as it is), the conversion block of x → θ that generates the angle information θref of the joint space based on the spatial trajectory Xref is also 1. Such robot operation is based on MOV
If EJ is named and the target joint angle and maximum speed are parameters, it can be described as a command MOVEJ (target joint angle, maximum speed). The speed profile is set by giving a trapezoidal rule or the like in advance.

2) World coordinate system / no external world sensor 5a use / no hand path designation In this case, the teaching point sequence in FIG. 7 or 8 is the target position / orientation, the velocity profile and the maximum velocity are given, and the Xref generator 7 is used. Is 1, and the block of x → θ converts the target position / orientation into the joint angle. The path of passage depends on the system. Such a robot operation can be described as a command of MOVEP (target position / posture, maximum speed) when MOVEP is named and the target position / posture and maximum speed are parameters.
The speed profile is set by giving a trapezoidal rule or the like in advance.

3) World coordinate system / outside world sensor 5a sensor not used / hand tip path specified In this case, an interpolation method between teaching point sequences and point sequences is given at the time of teaching, and the number of control points, control point array, and interpolation mode ( Specify the route from (Route generation mode). As the interpolation method, polygonal line interpolation, spline interpolation using a teaching point as a control point, Bezier curve interpolation, or the like is used, and CAD information or the like can be captured. Such a robot operation is called MOVE, and has the number of control points, the control point array, and the route generation mode described above as parameters, and the velocity profile and the maximum speed as parameters, MOVE (number of control points, control point array,
Route generation mode, maximum speed, speed profile) command. The velocity profile is also given here explicitly.

4) Use of world coordinate system / outside world sensor 5a In this case, the same method as 3) is used as a method of generating a teaching route from the teaching point sequence information from the teaching device 1 and a route designation method. The manipulator 4 together with additional information for processing the external sensor 5a for simplifying the processing algorithm.
A method of giving a control mode at the time of teaching for reflecting the information obtained from the external sensor 5a in the speed profile section 74 for the purpose of correcting the movement amount of is newly introduced.

In the operation of such a robot, TRACK (number of control points, arrangement of control points, route generation mode, additional information, maximum speed, speed profile) , Control mode). As shown in FIG.
The control of the control model is realized. Here, in the parameters, a) the number of control points, control point array, and route generation mode are used when the teaching route is specified by the teaching device, and b) additional information is used when the external device is sensed by the sensing device 5. C) Maximum speed, speed profile is manipulator 4
Is used to calculate the movement amount of
d) The control mode is used to obtain the correction amount of the movement amount of the manipulator 4 from the contact amount obtained by the contact sensing.

(Example of Method) Next, with reference to the drawings, a specific robot control method using the robot system apparatus α of the above-described apparatus example will be described. FIG. 9 shows a flowchart of the robot control method of this method example.

First, initialization is performed to set t = 0 and i = 0 (ST1). Next, the information of the teaching point i which is taught by the teaching device 1 and is stored in the memory (not shown) is read (S
T2). Then, additional information is selected from the read information and transmitted to the feature point processing unit 6 (ST3).

The additional information is, for example, information for selecting a feature point detection algorithm or information for designating a feature point to be detected. Here, the information for selecting the feature point detection algorithm is information that specifies which algorithm the teacher uses at the time of teaching when a plurality of pieces of feature point detection algorithm information are available. Further, the information designating the feature point to be detected is the information designating which feature point to select when the feature point detection algorithm detects a plurality of feature points or when there is a possibility of detection. The additional information is updated when the hand 4a of the manipulator 4 passes the control point.

The feature point processing section 6 selects the feature point detection algorithms I to III (ST4), and detects the feature points by the optimum detection algorithms I to III (S).
T5), the detection information is transmitted to the robot controller 2.

Next, processing of t = t + Δt is performed to add elapsed time to t (ST6). Then, from the sensing transmission information from the feature point processing unit 6 and the like, Xre of the robot controller 2 is
The f generator 7 calculates the attitude of the manipulator at the target position of the space curve Xref at time t (ST7).

Subsequently, the target position / orientation information Xref is transmitted from the robot control device 2 to the servo drive device 3 (ST
8). The servo control device 2 receives the target position / orientation information Xref and drives the manipulator 4 (ST9).

The procedure from step ST6 to step ST9 is executed until the manipulator 4 reaches the teaching point i, and when the manipulator reaches the teaching point i, the process proceeds to the next step ST11 (ST10).

In step ST11, it is confirmed whether the manipulator 4 has moved to the last teaching point. If it is determined that the last teaching point has not been reached, the process proceeds to step ST12, and if it is determined that the last teaching point has been reached, the processing is performed. To finish.

At step ST12, an addition process of i = i + 1 is performed, and the process returns to step ST2.

Next, the main processing contents of the robot control method of the present invention will be described in detail. (Path length display and path-to-orbit conversion processing) Assuming that the space curve is x (u) [a≤u≤b], and the length of the curve from the start point of x (u) is S,

[0066]

[Equation 2]

The following relational expression is obtained. Therefore, by giving a velocity profile in the form of v = f (s), f (s, ds / dt, ds2 / dt2) = 0, etc., the trajectory x (s) is changed to the trajectory x (s
(t)) is obtained. However, when the stop operation is involved, the time is explicitly specified for the stop period.

(Route Correction Method (Mapping Method)) In the route correction method, the target route Sw ′ (s) is generated by the above-mentioned equation (2).
In the robot system device α of this device example, as shown in FIG. 10, the correspondence between the teaching path Tw (s) and the sensing path Sw (s) is to be linked by any of parallel movement, congruential transformation, similarity transformation, and affine transformation. To do.

In this example of the apparatus, the sensing apparatus 5 always performs sensing and performs the robot work on the sensing path Sw (s), while the teaching path Tw (s) is used only for the section where the sensing fails. Target route S corrected to
A control method is adopted in which robot work is performed on w '(s), and when sensing is restored, robot work is performed on the sensing path Sw (s). However, when the work route of the robot system device α is short, after the sensing device 5 fails in sensing, the robot moves to the end with the target route Sw ′ (s) corrected based on the teaching route Tw (s). A control method for performing work may be used.

A specific processing method of the path correction method will be described below. 1) The starting point O ′ of the sensing path Sw (s) initially drives the manipulator in accordance with the teaching path Tw (s) stored in the memory of the robot controller 2, and after at least the sensing raw information stabilizes and stabilizes in the initial stage. , The first sensing point of the sensing device 5. In addition,
As a factor causing an error between the starting point O of the teaching path Tw (s) and the starting point O ′ of the sensing path Sw (s), (1) individual difference, (2)
Installation error and (3) sensing error can be mentioned.

2) The curve shape of the target path Sw '(s) created mainly by the sensing path Sw (s) is geometrically similar to the curve shape of the teaching path Tw (s) given by the teaching. Assuming that, the curve information of the teaching path Tw (s) is applied to the information of the sensing row to obtain the information.

3) In performing the mapping process, the initial value of the conversion formula (1) is determined. Sw (s) = Tw (s) + B (1) Here, the rotation matrix A is the unit matrix, and the translation vector B is the starting point O of the teaching path Tw (s) and the sensing path S.
The difference is the starting point O'of w (s). That is, S'w (s) = Tw (s) + B

4) The conversion formula (1) of the mapping process is updated as follows according to the contents of the mapping. In the case of parallel movement, A is a unit matrix and B is calculated from the path length S. In the case of the congruential transformation, the rotation matrix A is determined from the tangential vector of the route in addition to the translation amount. In the case of similarity transformation, the similarity ratio is determined from the curvature of the path, and the translation vector and rotation matrix are obtained. -In the case of affine transformation, the rotation matrix and translation vector are determined from the information of multiple points near the current position.

5) To detect the end point of the route, the following method is adopted.・ The distance between the world coordinates of the sensed point and the end point of the teaching,
Considering the measurement error, it is calculated from the sensing information of several neighboring points, and the point with the minimum distance is set as the end point. -Calculate the end point from the path length obtained from the taught path, and use this as the end point during playback. -The teaching route is converted by the conversion matrix to obtain the world coordinates of the end point, and compared with the world coordinates obtained by sensing, the point with the minimum distance is set as the end point.

By adopting the apparatus example and the method example of the present invention described above, the robot controller can control the target position / orientation of the hand of the manipulator moment by moment, and the precision of the control of copying the curved shape can be made high and the precision can be improved. Robot control is possible.

In particular, even if the characteristic point cannot be detected from the raw information measured by the external sensor 5a and the sensing fails, the manipulator 4 can detect the sensing information given up to that point and the teaching information previously taught. The position and orientation to be taken can be determined.

Although the typical apparatus examples and method examples of the present invention have been described above, the present invention is not necessarily limited to the means and method examples of these apparatus examples.
The present invention can be appropriately modified and carried out within a range where the object of the present invention is achieved and the effects described later are obtained. In particular, the method of designating the time and the speed of the hand of the manipulator 4 at that time was adopted to create the speed profile, but the hand speed of the manipulator 4 at the point of the length of the path and the length on the path is specified. The method may be adopted.

[0078]

As described above, according to the present invention, the external sensor system and the servo control system are interconnected to create a target path, so that the manipulator can be driven on the work path with extremely high accuracy. Has the effect.

In particular, even when the external sensor falls into a state in which measurement is impossible, the manipulator can be moved to the target path, and in the copying control, the manipulator can be finely controlled even if it is roughly taught.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a robot system device as an example of a device according to the present invention.

FIG. 2 is a schematic view of the robot system device showing the mounting position of the external sensor.

FIG. 3 is a block diagram showing a main part of a robot controller of an apparatus example according to the present invention.

FIG. 4 is an explanatory view showing a curved shape generated on the two steel plate mating surfaces of the work object A;

FIG. 5 is a graph showing a result of measuring the ab cross section of FIG. 4.

FIG. 6 is a diagram showing an example of a velocity profile used in the apparatus example and the method example of the present invention.

FIG. 7 is a block diagram showing a first robot control model used in an apparatus example and a method example of the present invention.

FIG. 8 is a block diagram showing a second robot control model used in the apparatus example and the method example of the present invention.

FIG. 9 is a flowchart showing a procedure of a robot control method used in a method example of the present invention.

FIG. 10 is an explanatory diagram showing a relationship between a sensing route and a teaching route.

FIG. 11 is a block diagram showing a configuration of a conventional robot system device.

FIG. 12 is a block diagram showing a robot control model of a conventional example.

[Explanation of symbols]

 α, β ... Robot system device A ... Work target B ... Mating plane boundary O ... Teaching start point O '... Sensing start point Tw (s) ... Teaching path Sw (s) ... Sensing path S'w (s) ... Target path DESCRIPTION OF SYMBOLS 1 ... Teaching device 2 ... Robot control device 3 ... Servo drive device 4 ... Manipulator 4a ... Minions 5 ... Sensing device 5a ... External sensor 6 ... Feature point processing unit 7 ... Xref generator 70 ... Teaching route generation unit 71 ... Sensing route generation Part 72 ... Target route generation unit 73 ... Orbit generation unit 74 ... Velocity profile 75 ... S generator

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Kenichi Arakawa 1-1-6 Uchisaiwaicho, Chiyoda-ku, Tokyo Nihon Telegraph and Telephone Corporation (72) Inventor Yukihisa Katayama 1-6-6 Uchisaiwaicho, Chiyoda-ku, Tokyo Nihon Telegraph and Telephone Corp. (72) Inventor Hirohisa Tezuka 1-1-6 Uchisaiwaicho, Chiyoda-ku, Tokyo Nihon Telegraph and Telephone Corp.

Claims (11)

[Claims] 1. When the robot control means creates trajectory information of a manipulator and controls the operation of a robot, sensing information and teaching information for a work object are simultaneously input in parallel, and the two pieces of information are associated with each other. While executing, preferentially adopt the sensing information, and when the sensing information is interrupted during the work, perform mapping processing to immediately take in the teaching information at that time, and carry out the work consistently and continuously. A robot control method characterized by: 2. When the robot control means creates trajectory information of the manipulator to control the operation of the robot, the manipulator, the servo drive means for driving the manipulator based on the entered trajectory information, and the manipulator are used. The teaching means for teaching the work, the sensing means for detecting the external world information, and the robot control means are provided, and the teaching path is generated from the discrete point sequence information related to the work target taught by the teaching means. In addition, a sensing route is generated from the discrete point sequence information related to the work object obtained by the sensing means, and the sensing route and the teaching route are mapped to generate a target route. The robot control method is characterized in that the trajectory information of the manipulator is generated from the. 3. The robot control according to claim 2, wherein the discrete point sequence information obtained from the teaching means is numerical information representing a space curve by giving a plurality of control points and an interpolation method. Method. 4. The mapping process includes a geometry such as parallel movement, congruential transformation, similarity transformation, affine transformation, etc., so that a sensing path is a target path at a sensing success point, and a sensing path is a teaching path at a sensing failure point. The robot control method according to claim 1, wherein the converted target path is used. 5. The robot control method according to claim 1, wherein the manipulator has a structure having a plurality of axes. 6. The generation of a sensing route is performed by a feature point process for detecting a feature point of a work target, and the detected feature point is used as a reference point when a route is created. 4. The robot control method according to 4 or 5. 7. When detecting the amount of movement of the manipulator in order to create the trajectory information of the manipulator, a speed profile is created based on the speed pattern and the maximum speed of the hand of the manipulator, and the speed profile and the elapsed time are used. The robot control method according to claim 2, wherein the current movement amount of the manipulator is determined. 8. A manipulator, and a servo drive device for driving the manipulator based on input trajectory information,
In a robot system including a teaching device that teaches work by a manipulator, a sensing device that detects external world information, and a robot controller, the robot controller is a discrete device that relates to a work target taught by the teaching device. Teaching path generating unit that generates a teaching route from specific point sequence information, a sensing route generating unit that generates a sensing route from discrete point sequence information related to a work object detected by the sensing device, and the sensing A robot system apparatus comprising: a target route generation unit that maps a route and the teaching route; and a trajectory generation unit that generates trajectory information of the manipulator from the target route. 9. The trajectory generation unit is connected to a velocity profile unit that creates a velocity profile based on the velocity pattern of the hand of the manipulator and the maximum velocity, and an S generator that generates the movement amount of the manipulator from the velocity profile and the elapsed time. 9. The robot system apparatus according to claim 8, wherein the robot system apparatus is configured to obtain the current movement amount information of the manipulator. 10. The robot system apparatus according to claim 8, wherein the external sensor of the sensing device is arranged in a state where the tip of the manipulator is offset. 11. The robot system apparatus according to claim 8, wherein the manipulator has a structure having a plurality of axes.