Abstract
This paper proposes a model-free control framework for the path planning of the rigid and soft robotic manipulator using an intelligent algorithm called Weighted Jacobian Rapidly-exploring Random Tree (WJRRT). The optimization approach is used to model the path planning problem, which is independent of the robotic model, and then used the WJRRT algorithm to solve it. WJRRT algorithm not only explores the cartesian space for the end-effector of the robotic manipulator randomly but also directs it towards the goal-position when required. It is robust enough to tackle the uncertainties in the manipulator and make the computation of path planning more efficient. WJRRT assigned a fitness value to each node of the tree. Based on the fitness values algorithm computes the final path, which is a trade-off between efficiency and safety of the path. The simulation results of two, three, and seven degrees of freedom (DOF) robotic manipulators are presented and compared with JT-RRT, Bi-RRT, and TB-RRT algorithms. Experimental results are verified using a soft manipulator made from flexible materials, i.e., polypropylene and polychloroprene. Their flexible structure makes their control complex and creates uncertainties in the model. The simulation and experimental results demonstrate that WJRRT can efficiently and accurately control the motion of manipulators.
Highlights
Soft robotic manipulators draw inspiration from animals like, arthropods, starfish, and snakes and attracts immense attention from researchers and engineers [1]–[4]
TWO-ARM ROBOTIC MANIPULATOR First, Weighted Jacobian Rapidly-exploring Random Tree (WJRRT) is tested on the two-arm robotic manipulator and compared the results with the JT-Rapidly-exploring Random Tree (RRT) [55], Bi-RRT [56], and TB-RRT [57] algorithms
It uses an intelligent framework that integrates the Jacobian controller with Rapidly-exploring Random Tree known as Weighted Jacobian Rapidly-exploring Random Tree (WJRRT)
Summary
Soft robotic manipulators draw inspiration from animals like, arthropods, starfish, and snakes and attracts immense attention from researchers and engineers [1]–[4]. The biased variant of the RRT algorithm [32] allows the rapid expansion of tree until it reaches the goal-node by giving more weight to those nodes and branches that are directed towards goal-node and increases the accuracy and efficacy As it can navigate through the complex space, it has already been employed in different path planning schemes [32]–[37]. The highlights of the proposed method are as follow: 1) A model-free control framework for the path planning of a soft robotic manipulator includes obstacle avoidance. 2) Proposed algorithm is a robust variant of RRT integrated with Jacobian-transpose and weighted feasible paths to design a path planning framework for a soft robotic manipulator while avoiding obstacles.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have