CN102289553B - Method and device for solving disassembly path of product parts - Google Patents

Method and device for solving disassembly path of product parts Download PDF

Info

Publication number
CN102289553B
CN102289553B CN 201110288747 CN201110288747A CN102289553B CN 102289553 B CN102289553 B CN 102289553B CN 201110288747 CN201110288747 CN 201110288747 CN 201110288747 A CN201110288747 A CN 201110288747A CN 102289553 B CN102289553 B CN 102289553B
Authority
CN
China
Prior art keywords
node
pose
path
solution
stage
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
Application number
CN 201110288747
Other languages
Chinese (zh)
Other versions
CN102289553A (en
Inventor
刘检华
宁汝新
尚炜
刘密
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Institute of Technology BIT
Original Assignee
Beijing Institute of Technology BIT
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing Institute of Technology BIT filed Critical Beijing Institute of Technology BIT
Priority to CN 201110288747 priority Critical patent/CN102289553B/en
Publication of CN102289553A publication Critical patent/CN102289553A/en
Application granted granted Critical
Publication of CN102289553B publication Critical patent/CN102289553B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Processing Or Creating Images (AREA)
  • Image Analysis (AREA)

Abstract

The invention provides a method and device for solving the disassembly path of product parts. The method comprises the following steps: solving the path of a first solving stage by taking the assembled posture of the parts as the first start posture of the first solving stage to obtain the first end posture of the first solving stage; solving a path of the second solving stage by taking the first end posture as the second start posture of the second solving stage to obtain the second end posture of the second solving stage; solving a path of the third solving stage by taking the second end posture as the third start posture of the third solving stage to obtain the third end posture of the third solving stage; and connecting the first path from the first start posture to the first end posture, the second path from the second start posture to the second end posture, and the third path from the third start posture to the third end posture in an end-to-end mode to obtain a global disassembly path of the parts. Through the invention, the product disassembly path in a complicated environment can be solved.

Description

Disassembly path method for solving and the device of product component
Technical field
The present invention relates to mechanical engineering field, refer to especially a kind of disassembly path method for solving and device of product component.
Background technology
It is that to keep away barrier or satisfy the operation needs be the parts assembly path search take security as leading indicator that purpose is carried out that the assembly path of product component is found the solution, simultaneously be also whether the checking Assembly sequence of products is reasonable, guarantee the important means of product assembling capacity.The assembly path solution technique be towards the assembling design (DFA) in an important gordian technique.The feasibility of assembly path is directly connected to the feasibility of assembling scheme, and in the assembling planning process of modern product, assembly path planning has become one of committed step.
In the assembling of the complex products such as aircraft, boats and ships, automobile, spacecraft,, complex structure many due to product component, the planning of its dismounting/assembly path is to carry out in a complex environment that is comprised of product component, clamping apparatus etc.Rely on experience and interactive mode to realize assembly path planning fully, require a great deal of time and very difficult realizing route optimization, even occur causing the part parts even can't assemble planning due to the limitation of interactive approach.Utilize PC Tools dismantle/the automatic planning of assembly path is a technology that wide application background is arranged.
At present domestic, for assembly path automatic calculation technology, certain research has been arranged abroad, but still mainly rest on the stage of theory study, relevant practical tool is less, its performance also haves much room for improvement.
Summary of the invention
The technical problem to be solved in the present invention is to provide a kind of disassembly path method for solving and device of product component, can realize improving the solution efficiency in detaching products path in complex environment, the detaching products path being found the solution.
For solving the problems of the technologies described above, embodiments of the invention provide a kind of disassembly path method for solving of product component, comprising:
The pose of assembling of product component is set;
The described pose that assembled is found the solution the first start bit appearance in stage as first, described the first path of finding the solution the stage is found the solution, obtain described product component and find the solution first of the stage described first and stop pose;
Stop pose as the second second initial pose of finding the solution the stage with described first, described the second path of finding the solution the stage is found the solution, obtain described product component and find the solution second of the stage described second and stop pose;
Stop pose as the 3rd the 3rd initial pose of finding the solution the stage with described second, the described the 3rd path of finding the solution the stage is found the solution, obtain described product component and find the solution the 3rd of the stage the described the 3rd and stop pose;
Described first start bit appearance is arrived the first path of described the first termination pose, the described second initial pose arrives described second and stops the second path of pose and the described the 3rd initial pose and arrive the described the 3rd Third Road footpath that stops pose and join end to end, and obtains the overall disassembly path of described parts.
Wherein, the described process that described the first path of finding the solution the stage is found the solution comprises:
The root node that described first start bit appearance is set to set;
Random one first spatial pose that produces, corresponding the first node of described the first spatial pose;
Obtain in tree from the second nearest node of described the first node;
Described parts are moved according to default the first step-length along described the second node to described the first node direction, the 3rd node corresponding to new pose after described the first step-length that obtains moving;
At described the 3rd node place, and the crucial interpolation point on from described the second node to the path of described the 3rd node carries out collision detection to described parts;
if described parts have collision at described the 3rd node place and described crucial interpolation point place and patch model, expand described the second node, if expand unsuccessfully, the step of returning to random generation one first spatial pose re-executes, if expand successfully, judge whether the distance whether described parts and described patch model do not exist other collision of dough sheet level or described the 3rd node to arrive root node corresponding to described first start bit appearance reaches the first default local disassembly path length, if, stop described first and find the solution finding the solution of stage, otherwise the step of returning to random generation one first spatial pose re-executes, if collisionless connects described the second node and described the 3rd node formed path as described the first path.
Wherein, the method for described collision detection comprises:
Obtain a test ray;
Described test ray is intersected test with described parts and with patch model that described parts carry out crash tests respectively;
Obtain the line segment that described test ray is arranged in described parts and described patch model simultaneously;
Repeatedly obtain different test rays and respectively described parts and described patch model are intersected test, obtain a plurality of line segments;
If in described a plurality of line segment, length greater than the pre-set ratio threshold value, thinks that described parts and described patch model have produced collision greater than the ratio of the sum of the line segment number of preset length threshold value and described a plurality of line segments, otherwise think collisionless.
Wherein, the step of described the second node of described expansion comprises:
The leaf node of searching described in described tree under the second node farthest is the 4th node;
Obtain successively first former generation's node of described leaf node, second former generation's node is until the quantity of the former generation's node that obtains reaches a preset value or obtained the root node of described tree;
Obtain the average pose node of described former generation's node of described preset value quantity, perhaps the average pose node of a plurality of former generation's nodes that comprise described first former generation's node, second former generation's node the root node from described leaf node to described tree;
Replace described the second node with described the 4th node, and definite a new direction is from described average pose node to described the 4th node.
Wherein, the described process that described the second path of finding the solution the stage is found the solution comprises:
The root node that the described second initial pose is set to set;
The random second space pose that produces, corresponding the 5th node of described second space pose;
Obtain in tree from the 6th nearest node of described the 5th node;
Obtain random number r in interval [0,1], with the Extending probability p on described random number r and described the 6th node relatively, if r<p expands described the 6th node; If r>=p, making under described the 6th node farthest leaf node replace described the 6th node expands, if expand unsuccessfully, the step of returning to random generation one second space pose re-executes, if expand successfully, whether the distance that described the 6th node that the new expansion of judgement obtains arrives root node corresponding to described first start bit appearance reaches the second default local disassembly path length, if reach, stop described first and find the solution the stage, otherwise the step of returning to random generation one second space pose re-executes.
Wherein, the process of described six nodes being expanded comprises:
The direction of described parts along described the 6th node to described the 5th node moved according to a fixed step size, the new pose node after this step-length of calculating motion;
To described parts at described new pose node place and the crucial interpolation point in the path from described the 6th node to described new pose node carry out collision detection, if collisionless connects described the 6th node and described new pose node; If there is collision, the Extending probability of expanding failed node is reduced, and determine an extended mode, calculate described new pose node according to the described extended mode of choosing, described the 6th node is expanded to described new pose node.
Wherein, described extended mode comprises: to the sampled point Directional Extension, arrive the Directional Extension of this node to father's node of this node, random choice direction expansion and/or the one dimension Directional Extension along the space.
Wherein, the expansion weights of described leaf node remain 1.
Wherein, the described step that the described the 3rd path of finding the solution the stage is found the solution is specially:
Adopt two-way Quick Extended random tree algorithm that the described the 3rd path of finding the solution the stage is found the solution.
Wherein, said method also comprises:
According to the global path step-length of setting, the path point sequence in described overall disassembly path is carried out interpolation and smooth.
Wherein, said method also comprises:
To the path point sequence inverted sequence on described overall disassembly path, obtain the assembly path of described parts.
Embodiments of the invention also provide a kind of disassembly path solving device of product component, comprising:
Module is set, is used for arranging the pose of assembling of product component;
First finds the solution module, is used for the described pose that assembled is found the solution the first start bit appearance in stage as first, and described the first path of finding the solution the stage is found the solution, and obtains described product component at the described first first termination pose of finding the solution the stage;
Second finds the solution module, is used for will described first stopping pose as the second second initial pose of finding the solution the stage, and described the second path of finding the solution the stage is found the solution, and obtains described product component at the described second second termination pose of finding the solution the stage;
The 3rd finds the solution module, is used for will described second stopping pose as the 3rd the 3rd initial pose of finding the solution the stage, and the described the 3rd path of finding the solution the stage is found the solution, and obtains described product component at the described the 3rd the 3rd termination pose of finding the solution the stage;
Overall situation disassembly path generation module, be used for described first start bit appearance is arrived the first path of described the first termination pose, the described second initial pose arrives described second and stops the second path of pose and the described the 3rd initial pose and arrive the described the 3rd Third Road footpath that stops pose and join end to end, and obtains the overall disassembly path of described parts.
The beneficial effect of technique scheme of the present invention is as follows:
in such scheme, finding the solution of overall disassembly path is decomposed into a plurality of stages (comprises that above-mentioned first finds the solution the stage, second finds the solution the stage and the 3rd finds the solution the stage) carry out, used the derivation algorithm that adapts to its feature each stage, can stablize fast the dismounting/assembly path in complex environment is found the solution, and the method be one general, detaching products in feasible complex environment/assembly path method for solving, take the digitizing geometric model of product as the basis, pass through respective algorithms, utilize computing machine that detaching products/assembly path is found the solution, can stablize that to find the solution fast environment unknown and with slype, dismounting/assembly path in the complex environments such as a large amount of barriers.
Description of drawings
Fig. 1 is the disassembly path method for solving schematic flow sheet of product component of the present invention;
Fig. 2 is in method shown in Figure 1, and parts begin to obtain the process schematic diagram of global path through the first path, the second path and Third Road footpath from assembling pose;
Fig. 3 is that first of method of the present invention is found the solution the expansion process schematic diagram based on the Quick Extended random tree algorithm of history guiding that adopts in phase process;
Fig. 4 is that first of method of the present invention is found the solution the collision algorithm schematic diagram that adopts in phase process;
Fig. 5 repeatedly carries out the schematic diagram of ray detection in collision algorithm shown in Figure 4;
Fig. 6 is that second of method of the present invention is found the solution in node extended mode in the adaptive Quick Extended random tree algorithm that adopts in phase process the schematic diagram to the sampled point Directional Extension;
Fig. 7 is that second of method of the present invention is found the solution in node extended mode in the adaptive Quick Extended random tree algorithm that adopts in phase process the schematic diagram to the father node of this node to the propagation direction expansion of this node;
Fig. 8 is the schematic diagram that second of method of the present invention is found the solution random choice direction expansion in node extended mode in the adaptive Quick Extended random tree algorithm that adopts in phase process;
Fig. 9 is that second of method of the present invention is found the solution in node extended mode in the adaptive Quick Extended random tree algorithm that adopts in phase process the schematic diagram of a certain dimension Directional Extension along the space.
Embodiment
For making the technical problem to be solved in the present invention, technical scheme and advantage clearer, be described in detail below in conjunction with the accompanying drawings and the specific embodiments.
As shown in Figure 1 and Figure 2, the disassembly path method for solving of a kind of product component of embodiments of the invention comprises:
Step 11 arranges the pose of assembling of product component;
Step 12 is found the solution the first start bit appearance in stage with the described pose that assembled as first, and described the first path of finding the solution the stage is found the solution, and obtains described product component and finds the solution first of the stage described first and stop pose;
Step 13 stops pose as the second second initial pose of finding the solution the stage with described first, and described the second path of finding the solution the stage is found the solution, and obtains described product component and finds the solution second of the stage described second and stop pose;
Step 14 stops pose as the 3rd the 3rd initial pose of finding the solution the stage with described second, and the described the 3rd path of finding the solution the stage is found the solution, and obtains described product component and finds the solution the 3rd of the stage the described the 3rd and stop pose;
Step 15, described first start bit appearance is arrived the first path of described the first termination pose, the described second initial pose arrives described second and stops the second path of pose and the described the 3rd initial pose and arrive the described the 3rd Third Road footpath that stops pose and join end to end, (as shown in Figure 2, parts 21 are through the first path a, the second path b and Third Road footpath c to obtain the overall disassembly path of described parts, path a, b, c joins end to end, and obtains the overall disassembly path of parts).
this embodiment is mainly used in complex environment in the product component assembling process, finding the solution of the dismounting/assembly path of product component, complex environment refers to because product structure is complicated, amount of parts is various, working environment has uncertain the grade and causes the obstacle environment that dismounting/assembly path is found the solution may have slype, intensive barrier, the difficult problems such as the easy change of scene, existing algorithm and instrument solution efficiency are low maybe can't be solved, this embodiment of the present invention is decomposed into a plurality of stages with finding the solution of overall disassembly path and (comprises that above-mentioned first finds the solution the stage, second finds the solution the stage and the 3rd finds the solution the stage) carry out, used the derivation algorithm that adapts to its feature each stage, can stablize fast the dismounting/assembly path in complex environment is found the solution, and the method be one general, detaching products in feasible complex environment/assembly path method for solving, take the digitizing geometric model of product as the basis, pass through respective algorithms, utilize computing machine that detaching products/assembly path is found the solution, can stablize that to find the solution fast environment unknown and with slype, dismounting/assembly path in the complex environments such as a large amount of barriers.
In another embodiment of the present invention, the Quick Extended random tree algorithm based on the history guiding that the process that described the first path of finding the solution the stage is found the solution adopts the present invention to propose, wherein, object collision in this algorithm detects collision detection and the appraisal procedure that adopts based on random ray test, wherein, the Quick Extended random tree algorithm based on the history guiding comprises:
The root node that described first start bit appearance is set to set;
Random one first spatial pose that produces, corresponding the first node of described the first spatial pose;
Obtain in tree from the second nearest node of described the first node;
Described parts are moved according to default the first step-length along described the second node to described the first node direction, the 3rd node corresponding to new pose after described the first step-length that obtains moving;
At described the 3rd node place, and the crucial interpolation point on from described the second node to the path of described the 3rd node carries out collision detection to described parts;
if described parts have collision (wherein at described the 3rd node place and described crucial interpolation point place and patch model, the patch model here can comprise: the patch model of other all parts and environmental model in current assembly environment, wherein, other parts are when being used as barrier, also can be called as patch model), expand described the second node, if expand unsuccessfully, the step of returning to random generation one first spatial pose re-executes, if expand successfully, judge whether the distance whether described parts and described patch model do not exist other collision of dough sheet level or the 3rd node to arrive root node corresponding to described first start bit appearance reaches the first default local disassembly path length, if, stop described first and find the solution finding the solution of stage, otherwise the step of returning to random generation one first spatial pose re-executes, if collisionless connects described the second node and described the 3rd node formed path as described the first path.
Wherein, the step of described the second node of described expansion comprises:
The leaf node of searching described in described tree under the second node farthest is the 4th node;
Obtain successively first former generation's node of described leaf node, second former generation's node, until the former generation's node quantity that obtains reach a preset value N (as N=8) or obtained as described in the root node of tree;
Obtain the average pose node of described former generation's node of described preset value quantity, perhaps the average pose node of a plurality of former generation's nodes that comprise described first former generation's node, described second former generation's node the root node from described leaf node to described tree;
Replace described the second node with described the 4th node, and definite a new direction is from described average pose node to described the 4th node.
The specific implementation process can be with reference to as follows:
Step 21 is treated the root node that the starting point of solution path (above-mentioned first finds the solution the stage) is set to set;
Step 22 produces a spatial pose q at random Rand(above-mentioned the first spatial pose);
Step 23 obtains in tree from the nearest node q of stochastic space pose Near(above-mentioned the second node);
Step 24 is along q NearTo q RandDirection is moved according to a fixed step size (wherein this step-length can preset), the new pose q after this step-length of calculating motion new(above-mentioned the 3rd node);
Step 25, to parts at new pose q newPlace and q NearTo q newCrucial interpolation point place, path carry out collision detection, if collisionless connects q NearAnd q newIf have collision, expand q Near, carry out following steps 251-254, return to net result;
Step 251 is sought q in tree NearFarthest leaf node q under node Front(above-mentioned the 4th node);
Step 252 is sought N former generation's node of leaf node successively, until the former generation's node quantity that obtains reach a preset value N (as N=8) or obtained as described in the root node of tree;
Step 253, the average posture information q of N former generation's node of calculating AverThe average pose node q of a plurality of former generation's nodes that comprise described first former generation's node, second former generation's node (average pose node) or the root node from described leaf node to described tree Aver
Step 254 determines that new expansion node is q Front, replace q NearNew propagation direction should be: from q AverTo q FrontExecution in step 24-25; As shown in Figure 3, q dirBe used in the drawings showing from node q FrontWith vector (q Front-q Aver) expand q for direction dir-q Front=q Front-q Aver
Step 26 is if the result of step 25 for to expand unsuccessfully, is returned to step 22; If expand successfully, judge new pose q newWhether whether the distance that arrives starting point (above-mentioned first start bit appearance) surpasses the first default local disassembly path length or parts and described patch model does not exist other collision of dough sheet level fully, if so, stops calculating; If not, return to step 22.
In above-mentioned Quick Extended random tree algorithm based on the history guiding, collision detection algorithm adopts collision detection and the appraisal procedure based on random ray test, and the implementation procedure of the method comprises:
Obtain a test ray;
Described test ray is intersected test with described parts and with patch model that described parts carry out crash tests respectively;
Obtain the line segment that described test ray is arranged in described parts and described patch model simultaneously;
Repeatedly obtain different test rays and respectively described parts and described patch model are intersected test, obtain a plurality of line segments;
If in described a plurality of line segment, length greater than the pre-set ratio threshold value, thinks that described parts and described patch model have produced collision greater than the ratio of the sum of the line segment number of preset length threshold value and described a plurality of line segments, otherwise think collisionless.
As shown in Figure 4 and Figure 5, should can be with reference to as follows based on the specific implementation process of the collision detection of random ray test and appraisal procedure:
If utilize the method that the collision situation of object A and object B is detected and assesses, the patch model that acquiescence has obtained A and B (wherein, this patch model refers to product component related in the solution procedure of path, clamping apparatus, working environment etc., wherein working environment comprises worktable, workshop etc.), the product component here, clamping apparatus and working environment etc. can be called under the prerequisite of how much patch model (being generally unordered tri patch model), and arthmetic statement is as follows:
Step 31, ratio threshold value 0<p<1 is set in preseting length threshold value t>0;
Step 32, random ray O of definition in the space 1D 1
Step 33 is done ray respectively and is intersected test with object A and object B, all dough sheets of traversal A and B in test, and the intersection point of recording ray and object in order are respectively and gather X OA={ A 1, A 2, A 3..., X OB={ B 1, B 2, B 3...; If ray and object A be without intersection point, or with object B without intersection point, return to step 32; As shown in Figure 4;
Step 34 is obtained respectively the line segment that ray is positioned at object A and object B inside, wherein, is arranged in line segment such as the A of object A 1A 2, A 2A 3And A 3A 4, be positioned at line segment such as the B of object B inside 1B 2If, set X OAElement number is even number, has S oA = { A j A j + 1 ‾ | A j , A j + 1 ∈ X OA , j = 1,3,5 , . . . , ( | X OA | - 1 ) } , If odd number has S oA = { A j A j + 1 ‾ | A j , A j + 1 ∈ X OA , j = 2,4,6 , . . . , ( | X OA | - 1 ) } . Same mode defines S OB
Step 35, traversal S OAAnd S OB, S oAB = { L = L OA ∩ L OB | ∀ L OA ∈ S OA , ∀ L OB ∈ S OB } , The line segment that namely is arranged in simultaneously object A and object B inside namely is positioned at the common factor of A and B, i.e. line segment B 1A 3And A 3B 2
Step 36 reflects according to the light reflection rule ray in the common factor of A and B, define new ray or redefine random ray (the random rays shown in a plurality of arrows as shown in Figure 5); Repeating step 33-36N time records testing result in S ALLIn;
Step 37, if the assessment and analysis testing result is S ALLMiddle length greater than p, thinks that A and B have produced degree of depth collision greater than the number of elements ratio of t; Otherwise think and produced shallow collision or collisionless.
In another embodiment of the present invention, the adaptive Quick Extended random tree algorithm that the process that described the second path of finding the solution the stage is found the solution adopts the present invention to propose, wherein, object collision in this algorithm detects and adopts traditional collision checking method, wherein, adaptive Quick Extended random tree algorithm comprises:
The root node that the described second initial pose is set to set;
The random second space pose that produces, corresponding the 5th node of described second space pose;
Obtain in tree from the 6th nearest node of described the 5th node;
Obtain random number r in interval [0,1], with the Extending probability p on described random number r and described the 6th node relatively, if r<p expands described the 6th node; If r>=p, making under described the 6th node farthest leaf node replace described the 6th node expands, if expand unsuccessfully, the step of returning to random generation one second space pose re-executes, if expand successfully, whether the distance that described the 6th node that the new expansion of judgement obtains arrives root node corresponding to described first start bit appearance reaches the second default local disassembly path length, if reach, stop described first and find the solution the stage, otherwise the step of returning to random generation one second space pose re-executes.Wherein, the second default local disassembly path length here can be identical with above-mentioned the first default local disassembly path length, also can be not identical.
Wherein, the process of described six nodes being expanded comprises:
The direction of described parts along described the 6th node to described the 5th node moved according to a fixed step size, the new pose node after this step-length of calculating motion;
To described parts at described new pose node place and the crucial interpolation point in the path from described the 6th node to described new pose node carry out collision detection, if collisionless connects described the 6th node and described new pose node; If there is collision, the Extending probability of expanding failed node is reduced, and determine an extended mode, calculate described new pose node according to the described extended mode of choosing, described the 6th node is expanded to described new pose node.
In this algorithm, described extended mode can comprise: to the sampled point Directional Extension, arrive the Directional Extension of this node to father's node of this node, random choice direction expansion and/or the one dimension Directional Extension along the space.
Preferably, the expansion weights of the described leaf node in this algorithm remain 1.
The specific implementation process of this self-adaptation Quick Extended random tree algorithm can be with reference to as follows:
Four kinds of node extended modes related in this algorithm comprise: to the sampled point Directional Extension; Father node to this node is expanded to the propagation direction of this node; Random choice direction expansion; The a certain dimension Directional Extension along the space is respectively as shown in Fig. 6,7,8,9.
Algorithm flow is described below:
Step 41 is treated the root node that the starting point of solution path is set to set, below repeating at the most to carry out for K time:
Step 42 produces a stochastic space pose q who is subjected to the terminal point guiding Rand(being above-mentioned second space pose);
Step 43 obtains in tree from the nearest node q of stochastic space pose Near(being above-mentioned the 6th node);
Step 44 is obtained random number r in interval [0,1], with q NearOn Extending probability p relatively, if r<p, to q NearExpand; If r>=p makes q NearUnder farthest leaf node q FrontReplace q NearExpand:
Step 45 is along q NearTo q RandDirection is moved according to a fixed step size, the new pose q after this step-length of calculating motion new
Step 46, to object at new pose q newPlace and q NearTo q newCrucial interpolation point place, path carry out collision detection, if collisionless connects q NearAnd q newIf there is collision, the Extending probability of expanding failed node reduced, and repeat at the most below N execution;
A chooses extended mode;
B calculates q according to extended mode new
C is with q NearTo q newExpansion;
D is if success returns results; If failure turns back to a;
Step 47 is if the result of step 44 for to expand unsuccessfully, is returned to step 42; If expand successfully, judge that whether close terminal point less than threshold value, if not, returns to step 42 to new node, if so, stops calculating.Wherein, in this expansion process, compose weights on each node, show by setting the weights size whether this point is easy to expansion (also claiming Extending probability), and control tree at the probability of this node growth; In the expansion of tree, if certain node is chosen as node to be expanded, further determine whether expand at this node according to the weights size; Preferably, the expansion weights of all leaf nodes always remain 1, guarantee that its Extending probability is not weakened; If expand unsuccessfully at a certain nonleaf node, reduce the weights of this node; If certain node is repeatedly expanded failure, it is 0 that its expansion weights are set, and closes the expansion of this node; When tree expansion runs into barrier, repeatedly attempt adopting different extended mode (as at above-mentioned a, b, c attempts in four kinds of extended modes of d) to expand, strengthen near the expansion of tree barrier; In different extended modes, select corresponding propagation direction, and with a fixed step size expansion.
In another embodiment of the present invention, the above-mentioned step that the described the 3rd path of finding the solution the stage is found the solution is specially: adopt traditional two-way Quick Extended random tree algorithm that the described the 3rd path of finding the solution the stage is found the solution.
Further, above-mentioned method shown in Figure 1 can also comprise:
According to the global path step-length of setting, the path point sequence in described overall disassembly path is carried out interpolation and smooth.
Further, above-mentioned method shown in Figure 1 can also comprise:
To the path point sequence inverted sequence on described overall disassembly path, obtain the assembly path of described parts.
To sum up, embodiments of the invention propose detaching products/assembly path method for solving in a kind of general, feasible complex environment, take the digitizing geometric model of product as the basis, by respective algorithms, utilize computing machine that detaching products/assembly path is found the solution.Principal feature of the present invention and advantage are: the proposition of (1) novelty based on Quick Extended random tree algorithm (RRT-H) algorithm, the self-adaptation Quick Extended random tree algorithm (RRT-A algorithm) of history guiding and collision detection and the appraisal procedure (RTSColDet algorithm) based on random ray test that adopts in the RRT-H algorithm, this RTSColDet algorithm can solve the conventional butt detection algorithm to the puppet of the collision detection between the two articles with face contact collision problem; The RRT-H algorithm can fast and stable be found the solution the path of simple motion in utmost point narrow space; The RRT-A algorithm can stablize that to find the solution fast environment unknown and with the dismounting/assembly path in the complex environments such as slype, a large amount of barriers.Method of the present invention product component, clamping apparatus, the working environment that the path solution procedure is related comprises all geometry patch model (being generally unordered tri patch model) input path solving devices such as worktable, workshop, and defines position and the attitude of each model in the space according to its posture information; Setting will be carried out the parts that find the solution in the path, it is set has assembled pose and unassembled pose, and path to be asked is to have assembled pose to the path of unassembled pose, and all the other all models are treated as the barrier in dismounting/assembling process; The step-length of global path is set; Path step-length in the RRT-H/RRT-A algorithm is set respectively; The collision depth threshold of RTSColDet is set; Then carry out the path solution procedure, this path solution procedure is found the solution the disassembly path of specified parts, and minute three phases realizes, first finds the solution the stage, second finds the solution the stage and the 3rd and find the solution the stage, calculates unsuccessfully as arbitrary stage, withdraws from this time and finds the solution; At first use the RRT-H algorithm to carry out finding the solution of the first path, wherein collision detection utilizes the RTSColDet algorithm to realize; First to find the solution the initial position and posture that RRT-H was set in the stage be parts assemblings pose in global path for this, do not set the termination position and posture of RRT-H; When the collision dough sheet number that returns when RTSColDet collision detection result is 0, again detect with the conventional butt detection algorithm, as collisionless, find the solution in the path that stops this stage; Otherwise whether the distance of the root node that the described phase one of node arrival of the new expansion of judgement sets reaches the first default local disassembly path length, and if so, find the solution in the path that stops this stage; This second stage of finding the solution uses RRT-A to carry out the path and finds the solution, use traditional collision detection algorithm to carry out collision detection; Set and (namely first find the solution the stage) on last stage the initial pose that the path termination pose of finding the solution is set as RRT-A, a certain node reaches the distance of setting apart from the global path starting point in the tree that RRT-A generates, and stops this and second finds the solution finding the solution of stage; The 3rd stage of finding the solution used RRT-ExtCon (two-way Quick Extended random tree algorithm) to carry out the path and finds the solution, use traditional collision detection algorithm to carry out collision detection; Set on last stage the path that (be above-mentioned second find the solution the stage) find the solution and stop the initial pose that pose is set as RRT-ExtCon, the termination pose of global path is the termination pose of RRT-ExtCon, stops the path behind the path that stops pose and finds the solution when solving this initial pose; Three sections paths that above-mentioned three stages of finding the solution are found the solution connect according to the order head and the tail, obtain the overall disassembly path from parts assembling pose to unassembled pose, are in fact a series of orderly pose path points; According to demand, as the needs assembly path, with this path point sequence inverted sequence; According to the global path step-length of setting, the path point sequence is carried out interpolation (being preferably linear interpolation), to obtain meticulousr smooth overall disassembly path.
In another aspect of this invention, embodiments of the invention also provide a kind of and disassembly path solving device (also can be called solver) the corresponding product component of said method, comprising:
Module is set, is used for arranging the pose of assembling of product component;
First finds the solution module, is used for the described pose that assembled is found the solution the first start bit appearance in stage as first, and described the first path of finding the solution the stage is found the solution, and obtains described product component at the described first first termination pose of finding the solution the stage;
Second finds the solution module, is used for will described first stopping pose as the second second initial pose of finding the solution the stage, and described the second path of finding the solution the stage is found the solution, and obtains described product component at the described second second termination pose of finding the solution the stage;
The 3rd finds the solution module, is used for will described second stopping pose as the 3rd the 3rd initial pose of finding the solution the stage, and the described the 3rd path of finding the solution the stage is found the solution, and obtains described product component at the described the 3rd the 3rd termination pose of finding the solution the stage;
Overall situation disassembly path generation module, be used for described first start bit appearance is arrived the first path of described the first termination pose, the described second initial pose arrives described second and stops the second path of pose and the described the 3rd initial pose and arrive the described the 3rd Third Road footpath that stops pose and join end to end, and obtains the overall disassembly path of described parts.
In said method, all realize that means and application scenarios all are applicable to also can reach identical technique effect in the embodiment of this device, do not repeat them here.
The above is the preferred embodiment of the present invention; should be pointed out that for those skilled in the art, under the prerequisite that does not break away from principle of the present invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.

Claims (10)

1. the disassembly path method for solving of a product component, is characterized in that, comprising:
The pose of assembling of product component is set;
The described pose that assembled is found the solution the first start bit appearance in stage as first, described the first path of finding the solution the stage is found the solution, obtain described product component and find the solution first of the stage described first and stop pose;
Stop pose as the second second initial pose of finding the solution the stage with described first, described the second path of finding the solution the stage is found the solution, obtain described product component and find the solution second of the stage described second and stop pose;
Stop pose as the 3rd the 3rd initial pose of finding the solution the stage with described second, the described the 3rd path of finding the solution the stage is found the solution, obtain described product component and find the solution the 3rd of the stage the described the 3rd and stop pose;
Described first start bit appearance is arrived the first path of described the first termination pose, the described second initial pose arrives described second and stops the second path of pose and the described the 3rd initial pose and arrive the described the 3rd Third Road footpath that stops pose and join end to end, and obtains the overall disassembly path of described parts;
The described process that described the first path of finding the solution the stage is found the solution comprises:
The root node that described first start bit appearance is set to set;
Random one first spatial pose that produces, corresponding the first node of described the first spatial pose;
Obtain in tree from the second nearest node of described the first node;
Described parts are moved according to default the first step-length along described the second node to described the first node direction, the 3rd node corresponding to new pose after described the first step-length that obtains moving;
At described the 3rd node place, and the crucial interpolation point on from described the second node to the path of described the 3rd node carries out collision detection to described parts;
if described parts have collision at described the 3rd node place and described crucial interpolation point place and patch model, expand described the second node, if expand unsuccessfully, the step of returning to random generation one first spatial pose re-executes, if expand successfully, judge whether the distance whether described parts and described patch model do not exist other collision of dough sheet level or described the 3rd node to arrive root node corresponding to described first start bit appearance reaches the first default local disassembly path length, if, stop described first and find the solution finding the solution of stage, otherwise the step of returning to random generation one first spatial pose re-executes, if collisionless connects described the second node and described the 3rd node formed path as described the first path.
2. the disassembly path method for solving of product component according to claim 1, is characterized in that, the method for described collision detection comprises:
Obtain a test ray;
Described test ray is intersected test with described parts and with patch model that described parts carry out crash tests respectively;
Obtain the line segment that described test ray is arranged in described parts and described patch model simultaneously;
Repeatedly obtain different test rays and respectively described parts and described patch model are intersected test, obtain a plurality of line segments;
If in described a plurality of line segment, length greater than the pre-set ratio threshold value, thinks that described parts and described patch model have produced collision greater than the ratio of the sum of the line segment number of preset length threshold value and described a plurality of line segments, otherwise think collisionless.
3. the disassembly path method for solving of product component according to claim 1, is characterized in that, the step of described the second node of described expansion comprises:
The leaf node of searching described in described tree under the second node farthest is the 4th node;
Obtain successively first former generation's node of described leaf node, second former generation's node is until the former generation's node quantity that obtains reaches a preset value or obtained the root node of described tree;
Obtain the average pose node of described former generation's node of described preset value quantity, perhaps the average pose node of a plurality of former generation's nodes that comprise described first former generation's node, second former generation's node the root node from described leaf node to described tree;
Replace described the second node with described the 4th node, and definite a new direction is from described average pose node to described the 4th node.
4. the disassembly path method for solving of product component according to claim 1, is characterized in that, the described process that described the second path of finding the solution the stage is found the solution comprises:
The root node that the described second initial pose is set to set;
The random second space pose that produces, corresponding the 5th node of described second space pose;
Obtain in tree from the 6th nearest node of described the 5th node;
Obtain random number r in interval [0,1], with the Extending probability p on described random number r and described the 6th node relatively, if r<p expands described the 6th node; If r 〉=p, making under described the 6th node farthest leaf node replace described the 6th node expands, if expand unsuccessfully, the step of returning to random generation one second space pose re-executes, if expand successfully, whether the distance that described the 6th node that the new expansion of judgement obtains arrives root node corresponding to described first start bit appearance reaches the second default local disassembly path length, if reach, stop described first and find the solution the stage, otherwise the step of returning to random generation one second space pose re-executes.
5. the disassembly path method for solving of product component according to claim 4, is characterized in that, the process that described six nodes are expanded comprises:
The direction of described parts along described the 6th node to described the 5th node moved according to a fixed step size, the new pose node after this step-length of calculating motion;
To described parts at described new pose node place and the crucial interpolation point in the path from described the 6th node to described new pose node carry out collision detection, if collisionless connects described the 6th node and described new pose node; If there is collision, the Extending probability of expanding failed node is reduced, and determine an extended mode, calculate described new pose node according to the described extended mode of choosing, described the 6th node is expanded to described new pose node.
6. the disassembly path method for solving of product component according to claim 5, it is characterized in that, described extended mode comprises: to the sampled point Directional Extension, arrive the Directional Extension of this node to father's node of this node, random choice direction expansion and/or the one dimension Directional Extension along the space.
7. the disassembly path method for solving of product component according to claim 5, is characterized in that, the expansion weights of described leaf node remain 1.
8. the disassembly path method for solving of product component according to claim 1, is characterized in that, the described step that the described the 3rd path of finding the solution the stage is found the solution is specially:
Adopt two-way Quick Extended random tree algorithm that the described the 3rd path of finding the solution the stage is found the solution.
9. the disassembly path method for solving of according to claim 1-8 described product components of any one, is characterized in that, also comprises:
According to the global path step-length of setting, the path point sequence in described overall disassembly path is carried out interpolation and smooth.
10. the disassembly path method for solving of product component according to claim 9, is characterized in that, also comprises:
To the path point sequence inverted sequence on described overall disassembly path, obtain the assembly path of described parts.
CN 201110288747 2011-09-26 2011-09-26 Method and device for solving disassembly path of product parts Expired - Fee Related CN102289553B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 201110288747 CN102289553B (en) 2011-09-26 2011-09-26 Method and device for solving disassembly path of product parts

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 201110288747 CN102289553B (en) 2011-09-26 2011-09-26 Method and device for solving disassembly path of product parts

Publications (2)

Publication Number Publication Date
CN102289553A CN102289553A (en) 2011-12-21
CN102289553B true CN102289553B (en) 2013-11-06

Family

ID=45335977

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 201110288747 Expired - Fee Related CN102289553B (en) 2011-09-26 2011-09-26 Method and device for solving disassembly path of product parts

Country Status (1)

Country Link
CN (1) CN102289553B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102708242B (en) * 2012-05-07 2014-08-06 上海飞机制造有限公司 Method and device for solving disassembling path of product pipe piece
CN106547936A (en) * 2015-09-22 2017-03-29 宫淼 A kind of method and system that dismounting path is obtained in aircraft maintainability is virtually verified
CN108959735B (en) * 2018-06-19 2021-07-13 浙江工业大学 Design method for matching assembly resources based on assembly availability model

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101477590A (en) * 2009-01-23 2009-07-08 清华大学 Method and system for mechanical and electrical disassembly planning and disassembly information management
CN101794334A (en) * 2010-02-11 2010-08-04 天津理工大学 Method for detaching products based on connected piece level network diagram

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4730440B2 (en) * 2009-01-01 2011-07-20 ソニー株式会社 Trajectory planning apparatus, trajectory planning method, and computer program
KR101554515B1 (en) * 2009-01-07 2015-09-21 삼성전자 주식회사 path planning apparatus of robot and method thereof
KR101667029B1 (en) * 2009-08-10 2016-10-17 삼성전자 주식회사 Method and apparatus of path planing for a robot

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101477590A (en) * 2009-01-23 2009-07-08 清华大学 Method and system for mechanical and electrical disassembly planning and disassembly information management
CN101794334A (en) * 2010-02-11 2010-08-04 天津理工大学 Method for detaching products based on connected piece level network diagram

Non-Patent Citations (16)

* Cited by examiner, † Cited by third party
Title
A Path Planning Approach to (Dis)Assembly Sequencing;Duc Thanh Le,et al.;《5th Annual IEEE Conference on Automation Science and Engineering》;20090825;全文 *
acute *
computers and information in engineering conference》.2011, *
Duc Thanh Le,et al..A Path Planning Approach to (Dis)Assembly Sequencing.《5th Annual IEEE Conference on Automation Science and Engineering》.2009,
es et al..Molecular Disassembly With Rrt-Like Algorithms.《IEEE International Conference on Robotics and Automation》.2007, *
es,et al..Disassembly Path Planning for Complex Articulated Objects.《IEEE TRANSACTIONS ON ROBOTICS》.2008,第24卷(第2期), *
Juan Cort&acute
Juan Cort&amp *
Liangjun Zhang,et al..D-Plan: Efficient Collision-Free Path Computation for Part Removal and Disassembly.《Computer-Aided Design and Applications》.2008, *
Wei SHANG,Jian-hua LIU.an adaptive rapidly-exploring random tree algorithm for assembly path planning in complex environments.《Proceeding of the ASME 2011 international design engineering technical conferences & computers and information in engineering conference》.2011,
Wei SHANG,Jian-hua LIU.an adaptive rapidly-exploring random tree algorithm for assembly path planning in complex environments.《Proceeding of the ASME 2011 international design engineering technical conferences &amp *
彭飞 等.旋转约束的船舶装配拆卸路径规划算法.《计算机仿真》.2011,第28卷(第6期),
彭飞,赵耀.基于约束通道的船舶装配拆卸双向砒汀路径规划.《中国造船》.2010,第51卷(第4期), *
旋转约束的船舶装配拆卸路径规划算法;彭飞 等;《计算机仿真》;20110630;第28卷(第6期);全文 *
机电产品拆卸序列生成研究;高建刚 等;《机械设计与研究》;20030228;第19卷(第1期);全文 *
高建刚 等.机电产品拆卸序列生成研究.《机械设计与研究》.2003,第19卷(第1期),

Also Published As

Publication number Publication date
CN102289553A (en) 2011-12-21

Similar Documents

Publication Publication Date Title
CN104155974A (en) Path planning method and apparatus for robot fast collision avoidance
US10124488B2 (en) Robot control system and method for planning driving path of robot
CN103529843B (en) Lambda path planning algorithm
EP2985122A1 (en) Method and apparatus for saving energy and reducing cycle time by optimal ordering of the industrial robotic path
CN103558512A (en) Method for locating power distribution network 10kV feeder line fault based on matrix operation
CN103632606B (en) Information processing method and device
CN102289553B (en) Method and device for solving disassembly path of product parts
CN104217251B (en) Equipment failure Bayesian network forecasting method based on K2 algorithms
CN105445624A (en) Cable fault positioning method according to combination of wavelet transformation and curve fitting
EP2998078A1 (en) Method for improving efficiency of industrial robotic energy consumption and cycle time by handling orientation at task location
CN102393646B (en) Multilayered dynamic collision detecting method and system for blade production line
Maddern et al. Towards persistent indoor appearance-based localization, mapping and navigation using cat-graph
CN106370059A (en) Random quick smooth second-order sliding mode terminal guidance method
Pipattanasomporn et al. Two-finger caging of concave polygon
CN106021647A (en) A cut sequence set-based dynamic fault tree Monte-Carlo simulation quantitative calculation method
CN104834809A (en) Artificial colony search-based seven-degree-of-freedom mechanical arm reverse kinematical solving method
CN102609321B (en) Rapid and continuous collision detection method based on K_DOPs (k-discrete orientation polytopes)
Marshall et al. Periodic formations of multivehicle systems
Mohamad et al. Ant colony robot motion planning
CN109840371A (en) A kind of dynamic multilayer coupling network construction method based on time series
Kantaros et al. Intermittent connectivity control in mobile robot networks
CN115793647A (en) Robot obstacle avoidance path planning method, system and medium
CN105590145A (en) Identification method and apparatus of shortest connecting distance between primary devices in substation
CN104679844A (en) Intermittent process batch data synchronizing method based on improved DTW (Dynamic Time Wrapping) algorithm
CN104331080A (en) Fixed point track route planning method for mobile robot

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C53 Correction of patent for invention or patent application
CB03 Change of inventor or designer information

Inventor after: Liu Jianhua

Inventor after: Ning Ruxin

Inventor after: Shang Wei

Inventor after: Liu Mi

Inventor before: Shang Wei

Inventor before: Liu Jianhua

Inventor before: Ning Ruxin

Inventor before: Liu Mi

COR Change of bibliographic data

Free format text: CORRECT: INVENTOR; FROM: SHANG WEI LIU JIANHUA NING RUXIN LIU MI TO: LIU JIANHUA NING RUXIN SHANG WEI LIU MI

C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20131106

Termination date: 20140926

EXPY Termination of patent right or utility model