Multitask-Based Anti-Collision Trajectory Planning of Redundant Manipulators

Area of Science:

• Robotics
• Artificial Intelligence
• Computational Geometry

Background:

• Redundant manipulators face challenges in multi-task execution due to obstacles.
• Task sequencing and joint trajectories are significantly impacted by environmental constraints.

Purpose of the Study:

• To develop an effective anti-collision trajectory planning method for redundant manipulators performing multiple tasks.
• To optimize the sequential order of task areas and joint trajectories while ensuring obstacle avoidance.

Main Methods:

• A special encoding genetic algorithm (SEGA) was proposed for multitask-based anti-collision trajectory planning.
• Spatial occupancy and obstacle detection were modeled using spherical enclosing box and superposition theories.
• Joint trajectories were represented by sixth-degree polynomials, with optimization of the sixth-order coefficient.

Main Results:

• The problem was transformed into a parameter optimization problem by optimizing task area sequence, collision detection, and joint trajectories.
• The SEGA cost function balances end-effector path length and joint angle variation.
• Numerical simulations validated the efficacy of the proposed SEGA method.

Conclusions:

• The proposed SEGA effectively addresses multitask-based anti-collision trajectory planning for redundant manipulators.
• The method ensures smooth joint movements and optimal path planning in the presence of obstacles.