Two robots such as adept-one and puma-762 are selected to pick and place parts in their workspace. Workcell-1 for driver assembly with Adept-one Robot 31 Figure 3.10 Workcell-1 for motor alternator assembly with Adept-one Robot 31 Figure 3.11.
Terms related to robot motion planning problem
Applications of path planning
Objective of the present work
Methodology
Outline of the thesis
Summary
LITERATURE REVIEW
- Overview
- Some important literatures related to the present work
- Path planning and assembly sequence generation
- Soft computing techniques for optimization
- Motion planning of robot with obstacle avoidance
- Summary
1997 Considered an approach to motion planning that incorporates visual servo constraints into the computation of the motion plans. Philip Chan [3] introduced the pattern-matching system for automatic assembly sequence(s) generation, considering that the subassembly problem represented as a connection could be solved similarly to the traveling salesman problem, and developed a method to reduce the number of questions using of the above method. Based on assembly by separation philosophy, a candidate list composed of feasible and reasonable separation operations that was derived from separation matrix to guide the sequence construction in the solution space expressed implicitly and which guaranteed the geometric feasibility of sequences. The state transition rule and Local and global update rules were also defined to ensure acquisition of the optimal solutions.
Positive feedback caused fast discovery of good solutions, distributed computations avoided premature convergence, and greedy heuristics helped to find acceptable solutions in the early stages of the search process. In this paper the author related this to the controller structure, the maximum available control and the size of the available uncertainty. Young [15] proposed to use the given robot geometry to generate geometric constraints on the robot workspace.
Herry Sutanto and Rajeev Sharma [17] considered an approach to motion planning that incorporates visual service constraints into the computation of motion plans. Since multiple paths are possible in the robot workspace between parts of an assembly product, choosing the best path following all these constraints is a critical factor.
GENERATION OF PATH SEQUENCES
Overview
- Constraint method : An assembly is called logically infeasible if the pair of subassemblies joined by the tasks is not connected in liaison graph, which is a connected graph
- Matrix method : The matrix method is used for the selection of the subassembly sequences of a product. The possible subassemblies are automatically detected by satisfying
Here {SRAN} is the set of all possible combinations between the N initial parts and SAR is the modified SRAN to satisfy the assembly rules that exist between the initial parts. SFE is the modified SAR to meet the assembly constraints while still meeting the assembly rules. The set of all SFE sequences, {SFE}, is the search space for the best or optimal sequence.
Connection diagrams show the connections between parts and connection sequences are similar to assembly sequences. This method uses a priority relationship between the parts, and the connecting sequences of the assembly are created after some different steps. The geometric model and the technological relationship between the product components are represented by means of three matrices, as explained below.
Interference matrix (Ak): Interference matrix is that square matrix of order '' where aij=1, if the element ei interferes with the element ej during the translation in the direction +k, otherwise aij =0. Contact matrix (Bk): The contact matrix Bk of a product formed by an' elements e1 ,e2,……,en, is that square matrix of order '' where bij=1, if the element ei is in contact with the element ej along the direction +k, otherwise at =0.
Product modeling for assembly sequence generation
- TASK DECOMPOSITION FOR ADEPT-ONE OR PUMA-762 ROBOT FOR WORK CELL-1 AND WORK CELL-2
- WORKCELL-1 WITH ADEPT-ONE ROBOT
- WORKCELL-2 WITH PUMA-762 ROBOT
- WORKCELL-3 WITH ADEPT-ONE AND PUMA-762 ROBOT
1 Shaft a Pick-rotate-orient-preme-place 2 Blade b Pick-rotate-premove-place 3 Nut c Pick-move-insert-place 4 Blade d Pick-rotate-premak-place 5 Nut e Pick-move-insert - the city. Adept-one Better fit 4 Blade d Pick-rotate-move-place Adept-one Better fit 5 Nut e Pick-rotate-move-insert-.
Summary
Overview
Uncertainty
Here is another example for finding the shortest path if there is uncertainty in the position, orientation, size, or shape of the polygon.
Problems on configuration space
Path sequences for building a driver for Adept-one or Puma-762 Product-3 robots. Automotive Alternator Assembly.
Path Sequences for workcell-3 Product-1 (Grinder Assembly)
Different paths for different robot
Summary
SOFT COMPUTING TECHNIQUES 5.1 Overview
- Ant Colony Technique
- Applying ACO to assembly sequence planning
- The solution
- ACO Algorithm
- Summary
Solution construction: Considering the problem-dependent heuristic information and path trace intensity, each ant chooses the next visited one to move according to probability. Track update: Evaluate the solution and deposit pheromone on the solution paths according to the quality of the solution to know if it is better or not solution. The possible tracks of the ants that join the nest and the food are represented by the possible disassembly sequences of the components which, conversely, represent the assembly sequences;.
The nest is represented by the first component of the sequence, and the food by the last component;. The concept of trace length (which should be minimized) replaces the concept of sequence quality (which should be maximized), evaluated in terms of the number of product orientation changes. Artificial ants iteratively travel through a loop involving a trip construction that depends on artificial pheromone trails and heuristic information.
The main idea of a modified algorithm is that the good tours are the positive feedback given by the ants through the pheromone update. The shorter the trip, the more pheromones are deposited on the selected path. This means that the path is more likely to be selected in subsequent iterations of the algorithm.
In this study, the pheromone is expressed as a 5n X 5n matrix because one of the Z directions is limited in study. One of the big advantages is that, the optimal solution satisfies all the composition constraints, objective function and also it is a part of stable solutions Ω~. Where m is the number of ants that find the iteration-best sequences and ∆τk( )i,j is the amount of pheromone ant k deposits on the arcs it visited.
During sequence construction, updating local pheromones encourages the exploration of alternative solutions, while updating global pheromones encourages the exploitation of the most promising solutions. Update the best set of each ant if the iteration set is the best found so far f. The soft computing technique performed in this chapter generates the stable motion planning sequence that satisfies all constraints and optimizes the stable sequences to obtain the best result.
RESULTS AND DISCUSSION
- Overview
- Path sequences
- Results and discussions
- Summary
An ant colony-based approach has been used to generate optimal stable assembly sequence and then the optimal road sequence. As the number of parts increases in assembly products, these methods provide more sequences and it is quite difficult to get the optimal sequence. Ultimately, the sequence generated in the algorithm is the optimal separation sequence for that product.
6 Pillar1 d Select-move-insert-site 15 Shell j Select-move-insert-site 7 Sensor i Select-move-insert-site 16 Plug g Select-rotate-move-insert-. In the assembly of a car alternator, the optimal sequence of paths is: C-B-E-F-D-G-H-I-J-K-L-M-N-A Table 6.3: Breakdown of the task for the optimal sequence of the car alternator assembly for work cell-1 and work cell-2 and work cell-3. An optimized path for the grinder assembly, driver assembly and automotive alternator assembly is developed using ACO techniques. The results are shown and described in the figures above.
CHAPTER-7
CONCLUSION AND FUTURE SCOPE
Conclusion
Since this work is related to product assembly, several products are chosen and then the assembly product is disassembled into different parts. The industrial robots are selected based on the tasks to be performed and the weight, shape and size of the parts to be processed.
Future Scope of work
Tasks are assigned to robots and motion planning is done and all the executable sequences are developed. 2] De Fazio and Whitney," An Integrated Computing Tool for Generating and Evaluating Assembly Lines for Mechanical Products", Robotics and Automation, IEEE Transactions, VOL. 3] Chan P., "Automatic assembly sequence generation by pattern matching", IEEE Transaction on Systems, Man and Cybanetics, 1991, pp.
4]de Mello.L.S.H., Sanderson.A.C., "A Correct and Complete Algorithm for Generating Mechanical Assembly Sequences", Robotics and Automation, IEEE Transactions, VOL. T & Maramam.R, "ASSEMBLY CHARACTERISTICS FOR MECHANICAL PRODUCT DATA", International Symposium on Assembly and Task Planning, IEEE, 1997, pp. M.," Ant Colony System: A Cooperative Learning Approach to the Traveling Salesman Problem", IEEE Transactions on Evolutionary Computation.
A," An Algorithm for Planning Collision-Pa Paths Among Polyhedral Obstacles, Communications the ACM, volume 22, October 1979, pp. R, "Practical Motion Planning with Visual Constraints", Symposium International on Assembly and Task Planning, Proceedings of the 1997 IEEE, pp. K., "Point-to-point trajectory planning of redundant flexible robot manipulators using genetic algorithms", Robotica (2002) volume 20, pp.
