Parametric path modeling for an optical automatic seam tracker and real time robotic control system

• US 4,965,499 A
• Filed: 12/22/1988
• Issued: 10/23/1990
• Est. Priority Date: 12/31/1987
• Status: Expired due to Fees

• Current Assignee: Westinghouse Electric Company LLC (Brookfield Asset Management Incorporated)
• Original Assignee: Westinghouse Electric Company LLC (Brookfield Asset Management Incorporated)
• Inventors: Ellison, James F., Taft, Jeffrey D.
• Primary Examiner(s): Shoop, Jr., William M.
• Assistant Examiner(s): Martin, David
• Application Number:
  US07/288,651
• Time in Patent Office: 670 Days
• Field of Search: 318/568-573, 318/574-577, 318/567, 219/124.1, 219/124.22, 219/124.32, 219/124.34, 901/9-10, 901/42, 901/47, 364/513, 364/167-169, 356/1-2, 356/367, 356/303-304
• US Class Current: 318/568.11
• CPC Class Codes: B23Q 35/128 Sensing by using optical means
• CPC Class Codes: B25J 9/1664 characterised by motion, path, trajectory planning
• CPC Class Codes: B25J 9/1684 Tracking a line or surface by means of sensors
• CPC Class Codes: G05B 19/42 Recording and playback systems, i.e. in which the programme is recorded from a cycle of operations, e.g. the cycle of operations being manually controlled, after which this record is played back on the same machine
• CPC Class Codes: G05B 19/425 Teaching successive positions by numerical control, i.e. commands being entered to control the positioning servo of the tool head or end effector
• CPC Class Codes: G05B 2219/36417 Programmed coarse position, fine position by alignment, follow line, path adaptive
• CPC Class Codes: G05B 2219/37572 Camera, tv, vision
• CPC Class Codes: G05B 2219/45104 Lasrobot, welding robot
• CPC Class Codes: G05B 2219/4701 Edge detector, project line, inclined camera detects discontinuity

Abstract

A control processor in a robotic system treats optically sensed locations along the path ahead of the tool, which are classified in relation to elapsed distances and identified in 3-D coordinates and 3-axes tool orientation, to apply in real time and under feedforward a control signal representing the error between anticipated tool position and anticipated sensed location on the path. In the process, the anticipated sensed location is obtained by reference to a model made ahead of time with the stored sensed locations grouped as a function of elapsed distances and having a common algebraic feature, such as the slope or the change of slope. A taught path for the robot has been recovered by the control processor also based on elapsed distances and control is related to actual tool position provided by the robot.

Claims

1. In a robot system including a robot controller for controlling an effector end along a seam in relation to a taught path and seam tracker control system, the seam tracker control system including a sensor system, for sensing the seam ahead of the effective end and for deriving data representing discrete successive sensed locations upon the seam, and a control processor responsive to said sensor system and to said robot controller for deriving a control signal representative of an error between effector end positioning and a matching sensed location, the effector end being controlled by said robot controller in response to said control signal to provide seam tracking with the effector end;

the combination of;

means responsive to said robot controller and associated with said control processor for successively calculating the distances between two consecutive sensed locations, for accumulating said distances to provide iteratively the elapsed distance down to the last sensed location and for providing orderly samples each representing such elapsed distance to identify the respective sensed locations upon the seam;

look-up table means for storing data representing coordinates derived by the sensor system for each of said sensed locations in relation to and in the order of said orderly samples;

the control processor including means for grouping said look-up table data in accordance with a common algebraic characteristic defining a seam model;

means within said control processor successively responsive to a present and a last position of said effector end for extrapolating therefrom to an anticipated position and for deriving an anticipated elapsed distance thereto;

means within said control process operating with said grouped look-up table data for correlating said anticipated elapsed distance of the effector end with a corresponding sample of said storing means to derive corresponding sample of said storing means to derive coordinates of an anticipated effector end location on the seam;

comparator means within said control processor for deriving a present control signal as said control signal and in relation to said corresponding coordinates; and

the control processor controlling the robot controller with said present control signal so as to effectuate feed forward control of the effector end toward a new position defined by said anticipated elapsed distance;

wherein the error effector end positioning and matching sensed seam location is applied cyclically as a first signal by the control processor to the robot controller and is compensated by the robot controller in response thereto through concurrent control of the effector end in relation to a taught path therein;

wherein the robot controller cyclically, and following said effector end cyclical control, provides the control processor with a first signal representative of such present position;

wherein the control processor cyclically generates with said first and control signals successive second signals representative of successive corresponding present positions characteristic of a recovered taught path;

wherein said extrapolating means provide with consecutive said second signals for a present and a last position characteristic on said recovered taught path, a third signal representative of an extrapolated position on said recovered taught path;

said seam model being responsive to said third signal for deriving a fourth signal representative of a corresponding seam location;

said comparating means being responsive to said third and fourth signals for deriving said present control signal.

6. A method of operating a robot system including:

a robot controller operating with a taught path;

a control processor for iteratively controlling an effector end with an error representing control signal in relation to a recovered path derived by the control processor from said robot controller;

a seam sensor vision having a look-ahead optical sensor for deriving successive 2-D light stripe images representing discrete successive transversed views of the seam as sensed at successive locations upon the seam a predetermined distance ahead of the effector end;

an image processor for treating each of said strip images in real time to derive in relation to each image a first signal representative of the coordinates for the respective sensed locations; and

a control processor responsive to a second signal indicative of the present effector end position and to said first signal for generating said control signal as an indication of an error between present effector end position and corresponding sensed location;

with the following steps being performed by said control processor;

storing and classifying said first signal coordinates for sensed locations obtained ahead of the effector end as a function of the elapsed distance since a first location sensed at the start of the seam sensing process;

said first signal coordinates following each other in the succession of sensed locations to form a discrete series of signals of different magnitudes;

grouping such discrete signals according to a common algebraic characteristic typical of a particular seam model;

storing such seam model grouped discrete signals as a function of the successively encountered sensed locations;

said control processor performing the following additional steps;

calculating and storing iteratively in response to said control signal and to said second signal the elapsed distance of the effector end since a start position thereof;

deriving iteratively with said second signal for a present effector end position and with said control signal for a present positioning operation of the effector end successive positions representing the taught path as recovered therefrom;

extrapolating by a predetermined amount from past positions on said recovered taught path to an anticipated position thereon;

extrapolating by said predetermined amount from past elapsed distances to an anticipated elapsed distances for said effector end;

determining with corresponding seam model grouped discrete signals in relation to said anticipated elapsed distance the coordinates of the corresponding sensed locations and deriving said first signal accordingly; and

controlling by feedforward the robot controller in response to the control signal derived from said first and second signals.

7. The method of claim 6, with said seam model grouped discrete signals defining within said seam model at least one of the slope and the change of slope as the common algebraic characteristic thereof.

8. A robotic control system including:

a robot controller operative to position an effector end in relation to a seam;

an image processor operative with an optical seam tracker and a look-ahead sensor mounted on said effector end a distance ahead therefrom to provide successive sensed locations on the seam; and

a control processor responsive to an instantaneous present position of the effector end and operative upon said robot processor for control of the effector end toward an anticipated position by feedforward in relation to a corresponding seam location derived from said image processor;

the image processor providing an indication of the successive sensed locations on the seam counted at respective elapsed distances from a starting point on the seam;

an indication of the successive effector end positions being derived from the robot processor and received by the control processor and counted at respective elapsed distances from said starting point;

wherein said corresponding seam location is ascertained by the control processor in relation to successive sensed locations derived from the image processor and to an elapsed distance corresponding to said anticipated position;

wherein the robot has a taught path and the control processor derives from the robot processor a recovered taught path, an extrapolated elapsed distance being ascertained on said recovered taught path by the control processor matching said anticipated position, and wherein the corresponding seam location is derived from said sensed locations as a function of said extrapolated elapsed distance;

said sensed locations being stored and grouped within the control processor in a function generator so as to form as a discrete function of elapsed distances a seam model;

said extrapolated elapsed distance being applied by the control processor to said functions generated as a variable representative signal so as to derive a signal representative of said corresponding location.

2. The system of claim 10 with said sensed seam locations being grouped in response to a predetermined minimum change of slope between consecutive derived sensed seam locations.

3. The system of claim 2 with said common incremental characteristic being derived by the control processor from successive sensed seam locations in accordance with the least-squared error line method.

4. The system of claim 2 with successive predetermined minimum changes of slope are detected as an indication of a circular seam path.

5. The system of claim 2 with said corresponding coordinates including a vector o representing the orientation of the effector end for the present position, the orientation vector o being successively determined by the control processor as a function of the elapsed distance and in relation to the same model grouped look-up table data, with the effector end anticipated position and orientation for control thereof being derived in accordance with said model common algebraic characteristic and for said anticipated elapsed distance.

9. The system of claim 8, with said function generator containing a plurality of said seam models formed successively by the control processor with the sensed locations derived by the seam sensor ahead of the effector end, a selection of an effective one of said seam models being effected by the control processor in relation to said extrapolated elapsed distance.

10. The system of claim 9, with a seam model being characterized by a linear function of said discrete elapsed distances representing the slope of the seam.

11. The system of claim 10 with said seam models defining successive changes of said seam slope from one another with the occurrence of said discrete distances in succession.

12. The system of claim 8, with feedforward control of the effector end being performed from a present position related to the recovered path by reference to the associated corresponding seam location, said extrapolated effector end position being determined from said present position and said anticipated elapsed distance being derived in relation thereto, said corresponding seam location defining an error used in the feedforward control in relation to said recovered taught path.

13. The system of claim 12, with such feedforward control being effected cyclically, and the elapsed distances being determined and stored by the control processor upon each new effector end position and extrapolated position.

14. The system of claim 13, with the look-ahead sensor being ahead of the effector end by a predetermined distance, with the succession of elapsed distances for the effector end successive positions and the succession of elapsed distances for the sensed seam locations being respectively displaced from one another by the amount of said predetermined distance.

15. A method of combining the operations of a seam sensor with the operations of a robotic system controlling an effector end in relation to the seam comprising the following steps:

sensing successive discrete locations on the seam ahead of the effector end while concurrently storing data representing seam coordinates at each of said seam locations as a function of the elapsed distances from the first of such seam locations;

modeling the stored data to generate a model representing the coordinates of said seam locations as a function of the elapsed distance;

deriving from the robotic system a representation of the effector end coordinates for successive discrete locations thereof;

calculating iteratively from said derived effector end coordinates an indication of the elapsed distance for each of the successive effector end locations and from the first of such effector end locations;

extrapolating from a present and a last effector end location an anticipated effector end location at an elapsed distance therefrom;

deriving from the stored data in said model the coordinates of the effector end at said anticipated effector end elapsed distance; and

controlling the robotic system in relation to said derived effector end coordinates.