CN104898470B - Robot teaching data generation system and method - Google Patents
Robot teaching data generation system and method Download PDFInfo
- Publication number
- CN104898470B CN104898470B CN201510151217.7A CN201510151217A CN104898470B CN 104898470 B CN104898470 B CN 104898470B CN 201510151217 A CN201510151217 A CN 201510151217A CN 104898470 B CN104898470 B CN 104898470B
- Authority
- CN
- China
- Prior art keywords
- training data
- sub
- output
- task
- processing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/04—Programme control other than numerical control, i.e. in sequence controllers or logic controllers
- G05B19/042—Programme control other than numerical control, i.e. in sequence controllers or logic controllers using digital processors
- G05B19/0426—Programming the control sequence
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Numerical Control (AREA)
- Manipulator (AREA)
Abstract
The present invention relates to robot teaching data generation system and method.Within the system, training data search part be not present in storage part with the training data of the processing of desired progress it is identical when, retrieve and whether there is and the sub- training data of operation identical first in processing again, if there is, then output it to training data editorial office, if there is no, then retrieve again with the presence or absence of the first sub- training data similar to the operation in the processing, if there is, then output it to training data correction portion, training data correction portion is based on the first sub- training data exported from training data search part, according to the operation to be performed, correct the first sub- training data, and it is output to training data editorial office, training data editorial office is based on the first sub- training data from training data search part or the output of training data correction portion, generation complete training data corresponding with the processing to be carried out.Thereby, it is possible to effectively reuse the training data generated.
Description
Technical field
The robot teaching data generation system of the training data of teaching is carried out to robot the present invention relates to generation and is shown
Teach data creation method.
Background technology
In the past, when generate robot training data when, as described in patent document 1 as it is necessary to have on machine
The professional knowledge of people's teaching and the technical staff of skilled operation degree carry out teaching action using the operating portion of teaching terminal etc.,
Thus training data is generated.Or, as described in patent document 2, technical staff for example, by PC etc. off-line teaching
Device, is built the threedimensional model of robot and the workpiece as object, robot teaching data is generated using the model.
Patent document 1：Chinese patent discloses CN103231374A；
Patent document 2：Japanese Patent Laid-Open 2012-091304.
The content of the invention
The prior art according to as such as patent document 1, patent document 2 is when generating training data, every time again
The new training data of generation, it is impossible to the part in training data or the training data that recycling has been generated.This
Sample, the load for generating system to training data is very big, and the burden of operating personnel is also very big.
The present invention provides a kind of robot teaching data generation system, including：Input unit, the input unit inputs teaching number
According to；Training data cutting part, one or more operations are decomposed into by the predetermined processing that robot to be performed, corresponding with each operation
Training data described in Ground Split generates the first sub- training data；Storage part, the storage part is accordingly stored for each operation
First sub- training data；Training data search part, when to generate new training data, first retrieves from storage part and whether there is
With the identical training data of training data of the processing of desired progress, if it is present by the training data as a result
Export to output section, if it does not exist, then retrieving of operation identical first that whether there is with the processing to be carried out again
Training data, if it is present output it to training data editorial office, if be still not present, then retrieves and whether there is
The first sub- training data similar to the operation in the processing to be carried out, is repaiied if it is present outputing it to training data
Positive portion, the operation for not retrieving the same or analogous first sub- training data notifies input unit input and the operation pair
The the first sub- training data answered；Training data correction portion, based on the first sub- training data exported from training data search part, root
According to the operation to be performed, the first sub- training data is corrected, and be output to training data editorial office；Training data editor
Portion, based on the first sub- training data from training data search part or the output of training data correction portion, generation is with to be carried out
The corresponding complete training data of processing, and be output to output section；And output section, output and the processing to be performed
Corresponding training data, and it is accordingly stored into storage part with operation.
In above-mentioned training data generation system, it is preferred that the training data cutting part is again by each breakdown of operation
For one or more tasks, and the first sub- training data generates the second sub- training data described in Ground Split corresponding with each task；Institute
State storage part and accordingly store the second sub- training data also directed to each task；The training data search part is not for retrieving
To the operation of the same or analogous first sub- training data, instead of notifying input unit to input the first sub- teaching corresponding with the operation
Data, and retrieve whether there is with the sub- training data of task identical second in the operation, if it is present outputing it to
Training data editorial office, if it does not exist, then retrieving again with the presence or absence of the second sub- teaching number similar to the task in the operation
According to if it is present training data correction portion is outputed it to, for not retrieving the same or analogous second sub- training data
Task, notify input unit to input corresponding with the task the second sub- training data；The training data correction portion also be based on from
Second sub- training data of training data search part output, according to being performed for task, corrects the second sub- training data；Institute
Training data editorial office is stated based on the described first sub- teaching number from training data search part or the output of training data correction portion
According to the second sub- training data, generation complete training data corresponding with the processing to be carried out, and be output to output
Portion；The output section exports training data corresponding with the processing to be performed, and it is accordingly stored with operation and task
To storage part.
In above-mentioned training data generation system, it is preferred that the input unit is teaching terminal or training data
Programmable device.
In above-mentioned training data generation system, it is preferred that described including training data generating means and server
Training data generating means and the server are connected in the way of it can communicate, wherein, the training data generating means bag
Include：The input unit, the training data cutting part, the training data search part, the training data correction portion, described show
Data edition portion, the output section are taught, the server includes the storage part.
In above-mentioned training data generation system, it is preferred that the storage part is cloud storage portion, the training data
Generating means and the server are connected by network cloud.
The present invention also provides a kind of training data generation method, including：Input step, inputs training data；Training data
Segmentation step, one or more operations are decomposed into by the predetermined processing that robot to be performed, Ground Split institute corresponding with each operation
State training data and generate the first sub- training data；Storing step, the first sub- training data is accordingly stored for each operation；Show
Data retrieval step, step is taught, when to generate new training data, first retrieval is with the presence or absence of the teaching number with the processing of desired progress
According to identical training data, if it is present the training data is exported to output section as a result, if it does not,
Then again retrieve with the presence or absence of with the sub- training data of operation identical first in the processing to be carried out, if it is present by its
Output, and be transferred to training data edit step, if be still not present, then retrieve with the presence or absence of with the processing to be carried out
Similar the first sub- training data of operation, if it is present outputting it, and training data amendment step is transferred to, for not having
There is the operation for retrieving the same or analogous first sub- training data, be transferred to input step input the first son corresponding with the operation
Training data；Training data amendment step, based on the first sub- training data exported in training data searching step, according to institute
The operation to be performed, corrects the first sub- training data, and be output to training data editorial office；Training data editor walks
Suddenly, based on the first sub- training data exported in training data searching step or training data amendment step, generation and institute
The corresponding complete training data of processing to be carried out, and output it, it is transferred to output step；And output step, output with
The corresponding training data of the processing to be performed, and it is accordingly stored with operation.
In above-mentioned robot teaching data creation method, it is preferred that in the training data segmentation step, then
It is one or more tasks by each breakdown of operation, and the first sub- of training data generation second described in Ground Split corresponding with each task
Training data；In the storing step, the second sub- training data is accordingly stored also directed to each task；In the teaching number
According in searching step, the operation for not retrieving the same or analogous first sub- training data, instead of being transferred to input step
Corresponding with the operation the first sub- training data of input, and retrieve whether there is and show with of task identical second in the operation
Data are taught, if it is present outputting it and being transferred to training data edit step, be whether there is if it does not exist, then retrieving again
The second sub- training data similar to the task in the operation, if it is present outputting it and being transferred to training data amendment step
Suddenly, for not retrieving the task of the same or analogous second sub- training data, it is transferred to input step input corresponding with the task
The second sub- training data；In the training data amendment step, also based on the output in the training data searching step
The second sub- training data, according to being performed for task, correct the second sub- training data；In training data editor step
In rapid, based on the described first sub- training data exported in training data searching step or training data amendment step and the
Two sub- training datas, generation complete training data corresponding with the processing to be carried out, output it and are transferred to output step；
In the output step, output training data corresponding with the processing to be performed, and by it with operation and task accordingly
Stored.
System and method are generated according to training data provided by the present invention, the teaching number generated can be reused
According to the especially partial data in training data, so as to mitigate the load that training data generates system, reduces operating personnel
Homework burden.
Brief description of the drawings
Fig. 1 is the block diagram that the training data schematically shown involved by the first embodiment of the present invention generates system；
Fig. 2 is the signal for showing training data generation system 1 and its relevant device involved by embodiments of the invention 1
Figure；
Fig. 3 is on soldering A as an example, and each operation and each first sub- training data are accordingly stored, incited somebody to action
The table that each task is accordingly stored with each second sub- training data；
Fig. 4 is the flow chart for an operating process for showing training data search part 104；
Fig. 5 is the flow chart for another operating process for showing training data search part 104；
Fig. 6 is on soldering B as an example, and each operation and each first sub- training data are accordingly stored, incited somebody to action
The table that each task is accordingly stored with each second sub- training data；
Fig. 7 is to show the training data generation system 1 involved by the variation of embodiments of the invention 1’And its correlation is set
Standby schematic diagram；
Fig. 8 is the signal for showing training data generation system 1 " and its relevant device involved by embodiments of the invention 2
Figure；
Fig. 9 is the flow chart for showing the training data generation method involved by second embodiment of the present invention；
Embodiment
First embodiment
As shown in figure 1, the robot teaching data generation system involved by the first embodiment of the present invention, including：It is defeated
Enter portion, training data cutting part, storage part, training data search part, training data correction portion, training data editorial office, output
Portion.
<Embodiment 1>
Fig. 2 is the signal for showing training data generation system 1 and its relevant device involved by embodiments of the invention 1
Figure.Relevant device mentioned here includes：Robot 100, robot controller 200 etc..
As shown in Fig. 2 the robot 100 is the industrial robot of welding robot, transfer robot etc..Robot
100 external structure mainly includes：Pedestal 10, large arm 20, forearm 30, wrist body 40, balancer 50 and multiple motors 60.
End effector (not shown) can be installed on the wrist body 40 of robot 100, performed according to the end effector different
Processing.
Robot controller 200 is used to enter robot 100 using the training data generated by training data generation system 1
Row control, makes it perform processing according to training data.
In embodiment 1, training data generation system 1 is realized by the microcomputers such as personal computer PC 2.
The microcomputer 2 includes：Input unit 101, training data cutting part 102, storage part 103, training data retrieval
Portion 104, training data correction portion 105, training data editorial office 106, output section 107.
Input unit 101 inputs training data.Here the training data inputted can be the teaching number for performing complete processing
According to or executable portion processing training data, that is to say, that can be the part in complete training data.By
In the existing training data of present invention recycling, therefore the input unit 101 is it can be appreciated that existing training data is generated
Equipment.For example, input unit 101 can include the composition such as input equipment, operational part of keyboard, the mouse of microcomputer 2 etc.
Training data editing machine.
Predetermined processing is decomposed into the one or more operations for realizing predetermined function by training data cutting part 102, and again
It is the realizing predetermined action of the task by each breakdown of operation.
In the present invention, predetermined processing can be divided into multiple operations sequentially in time, wherein, each operation is used for
Realize certain significant function.
Preferably, then by each operation one or more tasks are divided into sequentially in time, wherein, each task is used
In realizing predetermined action.
By taking welding robot as an example, each operation and task can be divided as follows.
For example, a soldering A can be divided into：Welding starts operation a, straight bead weld job b, circular arc weldering
Seam welding operation c, welding terminate operation d.
Preferably, each operation is divided into each task as follows again.
Welding is started into operation a to be divided into：Standard point location tasks a1, mobile task a2.
Straight bead weld job b is divided into：It is starting point location tasks b1, terminal location tasks b2, mobile task b3, straight
Wire bonding task b4.
Arch welded joint weld job c is divided into：Intermediate point location tasks c1, starting point location tasks c2, terminal positioning are appointed
Be engaged in c3, mobile task c4, circular arc weld task c5.
Welding is terminated into operation d to be divided into：Termination condition judges task d1, mobile task d2.
Training data described in the Ground Split also corresponding with each operation of training data cutting part 102 generates the first sub- training data,
The sub- training data of Ground Split each first corresponding with each task generates the second sub- training data.
As shown in figure 3, storage part 103 is by each operation and the first sub- training data and each task and the second sub- teaching number
According to accordingly being stored.
Training data search part 104 is retrieved when to generate new training data to storage part.
Specifically, training data search part 104 carries out following search operaqtion.
As shown in figure 4, first, in S1041 steps, retrieval whether there is and the processing of desired progress from storage part
The identical training datas of training data D2, if it is present the training data is exported to output section as a result, if
It is not present, then is transferred to S1042 steps.
Herein, function, the job title for constituting the processing, the correlation of processing carried out can be for example wanted by input
The mode of parameter value etc. is retrieved, and whether has the training data of matching in storage part.
In S1042 steps, then retrieve with the presence or absence of and the sub- teaching of operation identical first in the processing to be carried out
Data, if it is present outputing it to training data editorial office, if be still not present, are transferred to S1043 steps.
In S1043 steps, then retrieve with the presence or absence of the first sub- teaching similar to the operation in the processing to be carried out
Data, if it is present output it to training data correction portion, if be still not present, be transferred to S1044 steps or
S1046 steps.
Similar operation mentioned here refers to：The each task for constituting the operation it is identical and be only its order it is different and/
Or the different operation of parameter value of each task.
As shown in figure 5, the present embodiment 1 can also include S1044 steps and S1045 steps.In addition, S1044 steps and
S1045 steps can also be omitted, for example, operation not to be further divided into the situation of task.
In S1044 steps, for not both being determined as identical or not being determined as similar operation, retrieval whether there is
With the sub- training data of task identical second in the operation, if it is present output it to training data editorial office, if
Still it is not present, then is transferred to S1045 steps.
In S1045 steps, then retrieve with the presence or absence of the second sub- training data similar to the task in the operation, if
In the presence of then outputing it to training data correction portion, if be still not present, be transferred to S1046 steps.
Similar task mentioned here refers to：The task that the purpose of action is identical but parameter value is different.
In S1046 steps, for not retrieving the operation of the same or analogous first sub- training data or having inspection
Rope notifies input unit to input the training data of the part to the task of the same or analogous second sub- training data, and its is defeated
Go out and give training data editorial office.
As described above, training data search part 104 retrieves similar operation or in S1045 in S1043
Similar task is retrieved, then exports the sub- training data of corresponding first or the second son to training data correction portion 105
Training data.
Training data correction portion 105 is according to the operation or task to be performed, to what is exported by training data search part 104
First sub- training data or the second sub- training data are modified, generation corresponding with the operation to be performed or task first
Sub- training data or the second sub- training data.
Amendment mentioned here can be the amendment of the order of adjustment task, the parameter value for changing task etc..
Revised first sub- training data or the second sub- training data are output to teaching by training data correction portion 105
Data edition portion.
Training data editorial office 106 is based on the institute from training data search part 104 or the output of training data correction portion 105
State the first sub- training data and/or the second sub- training data, generation complete training data corresponding with the processing to be carried out
D2, and it is output to output section 107.
Output section 107 exports training data D2 corresponding with the processing to be performed, then, as shown in Figure 3, will
Training data D2 is accordingly stored in storage part with operation and/or task.
Hereinafter, followed by the example of above-mentioned welding robot, training data search part 104, teaching number are illustrated
According to correction portion 105, the operation of training data editorial office 106.
As shown in Figure 6, it is assumed that the soldering B that wanting the training data D2 of generation is used to realize includes：Welding starts to make
Industry a, right angle welding seam weld job e, welding terminate operation d '.
Also, welding, which starts operation a, to be included：Standard point location tasks a1, mobile task a2.
Right angle welding seam weld job e includes：Right angle electrical location tasks e1, starting point location tasks c2, terminal location tasks e3,
Mobile task c4 ', right angle welding task e5.
Welding, which terminates operation d ', to be included：Termination condition judges task d1 ', mobile task d2.
Training data search part 104 is proceeded as follows：
First, in S1041 steps, it is judged as without completely the same training data.
Then, in S1042 steps, it is judged as having identical welding to start operation a, a pairs of operation will be started with the welding
The the first sub- training data Da answered is exported to training data editorial office 106.
In S1043 steps, it is judged as that there is the welding similar to welding end operation d ' terminates operation d, will be with welding
Terminate the corresponding first sub- training data Dd of operation d to export to training data correction portion 105.
Training data correction portion 105 is based on being somebody's turn to do first sub- training data Dd corresponding with welding end operation d, by repairing
Just, generation first sub- training data Dd ' corresponding with welding end operation d ', and output it to training data editorial office 106.
In S1044 steps, it is judged as existing and appoints with the starting point location tasks c2 identicals in arch welded joint weld job c
Business, second sub- training data Dc2 corresponding with starting point location tasks c2 is exported to training data editorial office 106.
In S1045 steps, it is judged as there is the task similar to the mobile task c4 in arch welded joint weld job c
C4 ', second sub- training data Dc4 corresponding with mobile task c4 is exported to training data correction portion 105.
Training data correction portion 105 is based on being somebody's turn to do second sub- training data Dc4 corresponding with mobile task c4, by amendment,
Generation second sub- training data Dc4 ' corresponding with mobile task c4 ', and output it to training data editorial office 106.
Training data editorial office 106 is based on the first sub- training data Da exported from training data search part 104, the second son
Training data Dc2 and the first sub- training data Dd ', the second sub- training data Dc4 ' exported from training data correction portion 105,
Generation complete training data D2 corresponding with the soldering B to be performed.
Output section 107 exports the training data D2 of the generation, also, as shown in fig. 6, training data D2 is welded with constituting
Each operation and task for meeting processing B are accordingly stored in storage part.
, can by using stored the first sub- training data and/or the second sub- training data according to embodiment 1
Mitigate the load that training data generates system, reduce the homework burden of operating personnel.
<Variation>
As shown in fig. 7, generating system 1 as the training data of the above embodiments 1’It can include：Personal computer PC etc.
The teaching terminal 3 of microcomputer 2 and teaching box etc..
In the variation, input unit 101 can be teaching terminal 3 or keyboard including microcomputer 2,
The training data editing machine of input equipment, the operational part of mouse etc. etc..That is, in the variation, both can be by micro-
Type computer 2 generates training data, can also input training data by teaching terminal 3.
In the variation, during storage part 103 can be storage device or microcomputer 2 in teaching box
Storage device or both.
<Embodiment 2>
In example 2, as shown in figure 8, training data generation system 1 " includes：Server 300 and multiple training datas
Generating means 400-1~400-n.
Server 300 includes：Storage part, communication unit etc..
Training data generating means 400-1~400-n includes：Input unit, training data cutting part, communication unit, teaching number
According to search part, training data correction portion, training data editorial office, output section etc..
In example 2, server 300 and multiple training data generating means 400-1~sides of the 400-n that can communicate
Formula is connected.
Training data generating means 400-1~400-n can be the microcomputer of personal computer etc..Multiple teaching numbers
It can both be located at same workplace according to generating means 400-1~400-n, can also be with robot 100-1~100-n, machine
Device people controller 200-1~200-n is arranged respectively at different workplaces together.
Server 300 is used to storing and managing the teaching number uploaded by multiple training data generating means 400-1~400-n
According to.
Server 300 can be configured to as the more than one storage device connected by network cloud and computing dress
The aggregate put.That is, the storage part in server 300 can use the cloud storage device of Large Copacity.
In example 2, by the training data storage that multiple training data generating means are inputted or are generated to center clothes
It is engaged in the mass storage device in device, when multiple training data generating means need to generate new training data, to center clothes
Storage device in business device is retrieved, based on retrieval result, the generation training data to be used, and by the newly-generated teaching
Data are then stored into the storage device in server.
According to embodiment 2, due to uploading training data from multiple training data generating means to server, therefore, not only
It is that can utilize the training data for once inputting or generating in this training data generating means, can also utilizes by service
The training data for once inputting or generating in other training data generating means of device connection, the data volume for retrieval is big,
The training data of matching can be readily available.
Second embodiment
Second embodiment of the present invention provides show corresponding with the training data generating means in first embodiment
Teach data creation method.
As shown in figure 9, the training data generation method involved by second embodiment includes：S101 steps~S107 steps
Suddenly.
In S101 steps, training data is inputted, the training data is used for predetermined to realize to robot progress teaching
Processing.
In S102 steps, predetermined processing is decomposed into the operation for realizing predetermined function, Ground Split corresponding with each operation
The training data generates the first sub- training data.
Can also be again the realizing predetermined action of the task by each breakdown of operation, and corresponding with each task in S102 steps
The first sub- training data generates the second sub- training data described in Ground Split.
In S103 steps, the first sub- training data is accordingly stored for each operation.
In S103 steps, the second sub- training data can also be accordingly stored for each task.
As shown in Figure 4 and Figure 5, the S104 steps include：S1041~S1043, S1046 steps or S1041~
S1046 steps.Repeat specification is omitted herein.
In S105 steps, based on from the first sub- training data exported in training data searching step S104, according to institute
The operation to be performed, corrects the first sub- training data, outputs it, and be transferred to training data edit step S106.
In S105 steps, the second sub- training data exported in training data searching step S104, root are also based on
According to being performed for task, the second sub- training data is corrected, is output it, and is transferred to training data edit step S106.
In S106 steps, based on the output in training data searching step S104 or training data amendment step S105
The first sub- training data and/or the second training data, generation complete training data corresponding with the processing to be carried out, and
Output it, be transferred to output step S107.
In S107 steps, output training data corresponding with the processing to be performed, and by itself and operation (and task)
Accordingly storage part is arrived in storage.
In the above-described embodiment, it is illustrated by taking welding robot as an example, however, the invention is not restricted to bonding machine
Device people, is readily applicable to the robot of other purposes.
In the above-described embodiment, various processing robot carried out were divided for operation, the two stages of task,
But not limited to this, can be the stage of three and its above, for example, task can be further divided into subtask.
More than, although it is illustrated with reference to the embodiment of the present invention, but these embodiments are not
For limiting the present invention, and it is only the example of the present invention.Those skilled in the art will know that can be to these specific embodiment parties
Formula carries out various modifications, combination and changed, the idea of the invention that is limited without departing from claim or its equivalent and
Scope, then should regard as comprising in the present invention.
Claims (5)
1. a kind of robot teaching data generation system, including：
Input unit, the input unit inputs training data；
Training data cutting part, one or more operations are decomposed into by the predetermined processing that robot to be performed, with each operation pair
Training data described in Ground Split is answered to generate the first sub- training data；
Storage part, the storage part accordingly stores the first sub- training data for each operation；
Training data search part, when to generate new training data, first retrieval whether there is and desired progress from storage part
Processing the identical training data of training data, if it is present the training data is exported to output as a result
Portion, if it does not exist, then retrieve again with the presence or absence of with the sub- training data of operation identical first in the processing to be carried out, such as
Fruit is present, then outputs it to training data editorial office, if be still not present, then retrieves with the presence or absence of with to be carried out
The first similar sub- training data of operation in processing, if it is present training data correction portion is outputed it to, for not having
The operation of the same or analogous first sub- training data is retrieved, notifies input unit to input the first sub- teaching corresponding with the operation
Data；
Training data correction portion, based on the first sub- training data exported from training data search part, according to the work to be performed
Industry, corrects the first sub- training data, and be output to training data editorial office；
Training data editorial office, based on the first sub- teaching number from training data search part or the output of training data correction portion
According to, generation complete training data corresponding with the processing to be carried out, and it is output to output section；And
Output section, output training data corresponding with the processing to be performed, and it is accordingly stored into storage part with operation,
Each breakdown of operation is again one or more tasks by the training data cutting part, and described in Ground Split corresponding with each task
First sub- training data generates the second sub- training data；
The storage part accordingly stores the second sub- training data also directed to each task；
The training data search part is directed to the operation for not retrieving the same or analogous first sub- training data, instead of notifying
Input unit inputs corresponding with the operation the first sub- training data, and retrieve with the presence or absence of with the task identical in the operation the
Two sub- training datas, if it is present outputing it to training data editorial office, whether there is if it does not exist, then retrieving again
The second sub- training data similar to the task in the operation, if it is present training data correction portion is outputed it to, for
The task of the same or analogous second sub- training data is not retrieved, notifies input unit to input the second son corresponding with the task and show
Teach data；
The training data correction portion is also based on the second sub- training data exported from training data search part, according to being performed
Task, correct the second sub- training data；
The training data editorial office is based on first son from training data search part or the output of training data correction portion
Training data and the second sub- training data, generation complete training data corresponding with the processing to be carried out, and output it
To output section；
The output section exports training data corresponding with the processing to be performed, and it is accordingly stored with operation and task
To storage part.
2. robot teaching data generation system as claimed in claim 1, it is characterised in that
The input unit is teaching terminal or training data programmable device.
3. robot teaching data generation system as claimed in claim 1, it is characterised in that
Including training data generating means and server, the training data generating means and the server are can communicate
Mode is connected,
Wherein, the training data generating means include：The input unit, the training data cutting part, the training data
Search part, the training data correction portion, the training data editorial office, the output section,
The server includes the storage part.
4. robot teaching data generation system as claimed in claim 3, it is characterised in that
The storage part is cloud storage portion, and the training data generating means and the server are connected by network cloud.
5. a kind of training data generation method, including：
Input step, inputs training data；
Training data segmentation step, one or more operations are decomposed into by the processing that robot to be performed, corresponding with each operation
Training data described in Ground Split generates the first sub- training data；
Storing step, the first sub- training data is accordingly stored for each operation；
Training data searching step, when to generate new training data, first retrieval whether there is and the processing of desired progress
The identical training data of training data, if it is present the training data is exported to output section as a result, if not
In the presence of, then again retrieve with the presence or absence of with the sub- training data of operation identical first in the processing to be carried out, if it is present
Output it, and be transferred to training data edit step, if be still not present, then retrieve with the presence or absence of and the place to be carried out
The first similar sub- training data of operation in reason, if it is present outputting it, and is transferred to training data amendment step, pin
Operation to not retrieving the same or analogous first sub- training data, is transferred to input step input corresponding with the operation the
One sub- training data；
Training data amendment step, based on the first sub- training data exported in training data searching step, according to being held
Capable operation, corrects the first sub- training data, and be output to training data editorial office；
Training data edit step, based on the first son exported in training data searching step or training data amendment step
Training data, generation complete training data corresponding with the processing to be carried out, and output it, it is transferred to output step；With
And
Step, output training data corresponding with the processing to be performed are exported, and it is accordingly stored with operation,
It is one or more tasks in the training data segmentation step, then by each breakdown of operation, and with each task accordingly
Split the described first sub- training data and generate the second sub- training data；
In the storing step, the second sub- training data is accordingly stored also directed to each task；
In the training data searching step, the operation for not retrieving the same or analogous first sub- training data,
The first sub- training data corresponding with the operation is inputted instead of being transferred to input step, and retrieves and whether there is and appointing in the operation
The business sub- training data of identical second, if it is present output it and be transferred to training data edit step, if it does not,
Then retrieve again with the presence or absence of the second sub- training data similar to the task in the operation, if it is present outputting it and turning
Enter training data amendment step, the task for not retrieving the same or analogous second sub- training data is transferred to input step
Input the second sub- training data corresponding with the task；
In the training data amendment step, also based on the second sub- teaching number exported in the training data searching step
According to according to being performed for task, correcting the second sub- training data；
In the training data edit step, based on the output in training data searching step or training data amendment step
The described first sub- training data and the second sub- training data, generation complete teaching number corresponding with the processing to be carried out
According to outputting it and be transferred to output step；
In the output step, output training data corresponding with the processing to be performed, and by itself and operation and task pair
Ground is answered to be stored.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510151217.7A CN104898470B (en) | 2015-04-01 | 2015-04-01 | Robot teaching data generation system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510151217.7A CN104898470B (en) | 2015-04-01 | 2015-04-01 | Robot teaching data generation system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104898470A CN104898470A (en) | 2015-09-09 |
| CN104898470B true CN104898470B (en) | 2017-09-15 |
Family
ID=54031188
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510151217.7A Expired - Fee Related CN104898470B (en) | 2015-04-01 | 2015-04-01 | Robot teaching data generation system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104898470B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6470326B2 (en) * | 2017-01-31 | 2019-02-13 | ファナック株式会社 | Control device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2009262279A (en) * | 2008-04-25 | 2009-11-12 | Nec Corp | Robot, robot program sharing system, robot program sharing method, and program |
| CN101100060B (en) * | 2006-07-04 | 2010-06-16 | 发那科株式会社 | Device and method for preparing robot program |
| CN103514148A (en) * | 2012-06-27 | 2014-01-15 | 株式会社神户制钢所 | Training data editing system and program |
| CN104070266A (en) * | 2013-03-28 | 2014-10-01 | 株式会社神户制钢所 | Welding seam information setting device, program, automatic teaching system and welding seam information setting method |
| CN104423372A (en) * | 2013-08-30 | 2015-03-18 | 株式会社神户制钢所 | Teaching data making system and procedure |
-
2015
- 2015-04-01 CN CN201510151217.7A patent/CN104898470B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101100060B (en) * | 2006-07-04 | 2010-06-16 | 发那科株式会社 | Device and method for preparing robot program |
| JP2009262279A (en) * | 2008-04-25 | 2009-11-12 | Nec Corp | Robot, robot program sharing system, robot program sharing method, and program |
| CN103514148A (en) * | 2012-06-27 | 2014-01-15 | 株式会社神户制钢所 | Training data editing system and program |
| CN104070266A (en) * | 2013-03-28 | 2014-10-01 | 株式会社神户制钢所 | Welding seam information setting device, program, automatic teaching system and welding seam information setting method |
| CN104423372A (en) * | 2013-08-30 | 2015-03-18 | 株式会社神户制钢所 | Teaching data making system and procedure |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104898470A (en) | 2015-09-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7035711B2 (en) | Machining system | |
| US11224930B2 (en) | Weld sequencer part and statistical limits analyzer | |
| EP3441175A1 (en) | Weld sequencer part inspector | |
| US20190340954A1 (en) | Portable user interface for a welding type system | |
| US11648674B2 (en) | System and method for robotic bin picking using advanced scanning techniques | |
| EP3263268B1 (en) | Offline teaching device | |
| CN111738540A (en) | System and method for providing part quality confidence | |
| US11534910B2 (en) | Control device for displaying a relationship between robot output and device input | |
| CN104898470B (en) | Robot teaching data generation system and method | |
| JP6560841B1 (en) | Control device, work robot, program, and control method | |
| CN104679806A (en) | Retrieval supporting system of welding conditions | |
| JP2022132506A (en) | Simulation method, simulation system, and data structure of module data | |
| US12090581B2 (en) | Welding sequence guidance using three dimensional models | |
| JP5278057B2 (en) | Industrial robot position teaching device, industrial robot motion teaching device, and program | |
| JP6577686B1 (en) | Control device, work robot, program, and control method | |
| US11568112B2 (en) | System design support apparatus and system design support method | |
| JP2011059801A (en) | Program creation/instruction device and method | |
| JP4243384B2 (en) | Welding condition setting method and system | |
| JP7349610B1 (en) | Welding machine, welding work support method, and welding work support system | |
| US11768483B2 (en) | Distributed weld monitoring system with job tracking | |
| JP2002205168A (en) | Device and method for controlling welding robot | |
| JP6546269B2 (en) | Error handling support device | |
| JP2009098709A (en) | Robot teaching system | |
| JP2021064077A (en) | Parameter setting device and parameter setting system for machine tool | |
| JPH08214457A (en) | Power system diagram preparing equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: 442000 Shiyan East Economic Development Zone, East Ring Road, Hubei, No. 109 (Hubei long song industry and Trade Co., Ltd. office building) Applicant after: HUBEI QITONG INTELLIGENT TECHNOLOGY CO., LTD. Address before: 442000 Shiyan City, Hubei province Maojian District East Ring Road No. 111 Applicant before: HUBEI QITONG INTELLIGENT TECHNOLOGY CO., LTD. |
|
| COR | Change of bibliographic data | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170915 Termination date: 20190401 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |