Two-degree-of-freedom manipulator path planning based on zeroing neural network

Li, Yan; Liu, Keping

doi:10.1051/matecconf/202030904005

Cited by 2 publications

(1 citation statement)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several works will focus on the shortest path problem of manipulator path planning of 2-Dof robot arm without taking into account the best way to reach a desired position [39]. In this case, the approach presented in this paper deals with several improves of the path planning, with respect to the short time, the best trajectory and the avoidance of singular positions.…”

Section: Resultsmentioning

confidence: 99%

Path Planning of the Manipulator Arm FANUC Based on Soft Computing Techniques

Benchikh¹,

Ahmed‐Foitih²,

Bouzgou³

2020

IREACO

View full text Add to dashboard Cite

In this paper, direct and inverse geometric models for a 6 degrees of freedom manipulator robot arm are developed, and a set of homogeneous matrices are generated by Denavit-Hartenberg formalism. Moreover, a path planning method based on soft computing techniques is presented, which consists of using the neural network to model the end-effector workspace, and then determining the optimal trajectory to reach a desired position. The optimization of the trajectory depends on the minimization of the cost function, defined by the sum of two energies. The first one is the collision penalty (E c) generated by each obstacle shape; the second one is the trajectory length penalty (E l). Four steps are considered; at first, the configuration of the robot workspace where any assumption is taken into account. The robot is assimilated to a particle that moves inside a small two-dimensional space, and the obstacle is a polygon. Second, the Multilayer Perceptron Neural Network (MLP neural network) is used for the robot workspace modeling. Model block inputs are robot's previous position values and the outputs are robot's following position values. Then, the proposed approach is applied to optimize the cost function and to determine the end-effector trajectory in case of obstacles. The obtained result is a smooth trajectory, which represents robot motion. Finally, the presented approach is validated on a FANUC robot arm in a virtual environment, where the simulation results show this approach efficiency.

show abstract

Section: Resultsmentioning

confidence: 99%