Path Planning for Multi-Joint Manipulator Based on the Decomposition of Configuration Space
Zhou Fang, Zhu Qidan, Wang Yuquan, Tong Wang
Abstract
Zhou Fang, Zhu Qidan, Wang Yuquan, Tong Wang
Abstract
The problem of collision-free path planning for six joints manipulator used for nuclear reactor repairing was considered, a new approach based on the decomposition of configuration space was presented in this paper. Since the collision with the environment occurs mainly on the three joints near the proximal joints of manipulator, the six dimensional configuration space was decomposed into two three low dimensional subspaces. Respectively, the discrete configurations were generated in each subspace for the local path searching, then a collision table was set up for the on-line path planning. In order to reduce the size of collision table and avoid the redundant collision checking, in the process of sampling configuration, we check the collision status of these points and avoid the collided configuration points to be generated in the collision table. In order to further improve the reasonability distribution of sampled points, we used the distance information with the obstacle to guide the generation of configurations. The simulation results showed that the approach not only reduced the length of searing path, but also improved the efficient of on-line planning.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The problem of collision-free path planning for six joints manipulator used for nuclear reactor repairing was considered, a new approach based on the decomposition of configuration space was presented in this paper. Since the collision with the environment occurs mainly on the three joints near the proximal joints of manipulator, the six dimensional configuration space was decomposed into two three low dimensional subspaces. Respectively, the discrete configurations were generated in each subspace for the local path searching, then a collision table was set up for the on-line path planning. In order to reduce the size of collision table and avoid the redundant collision checking, in the process of sampling configuration, we check the collision status of these points and avoid the collided configuration points to be generated in the collision table. In order to further improve the reasonability distribution of sampled points, we used the distance information with the obstacle to guide the generation of configurations. The simulation results showed that the approach not only reduced the length of searing path, but also improved the efficient of on-line planning.
Key concepts: Collision, Configuration space, Path (computing), Motion planning, Linear subspace, Computer science, Subspace topology, Collision detection