Obstacle-avoidance trajectory planning method for excavators in confined spaces based on improved RRT* algorithm

Area of Science:

• Robotics and Automation
• Construction Engineering
• Artificial Intelligence

Background:

• Construction machinery is transitioning to unmanned autonomous operation.
• Excavators require sophisticated trajectory planning for complex and confined environments.
• Existing algorithms may lack efficiency and smoothness for autonomous tasks.

Purpose of the Study:

• To propose an obstacle avoidance trajectory planning scheme for autonomous excavators.
• To enhance path search efficiency and trajectory quality in restricted conditions.
• To achieve multi-objective optimization for operation duration and motion smoothness.

Main Methods:

• Developed a simulation model for excavator mechanisms and environments.
• Enhanced the RRT* algorithm with environmental parameter-based heuristic search and adaptive goal-biased strategy.
• Utilized quintic Non-Uniform Rational B-Spline (NURBS) curves for bucket tip trajectory optimization.

Main Results:

• Reduced path length by 3.65%, iteration count by 64.15%, and computation time by 67.9%.
• Improved trajectory smoothness by 33.4%.
• Generated collision-free, smooth, and energy-efficient trajectories.

Conclusions:

• The enhanced RRT* algorithm significantly improves autonomous excavator performance.
• The approach ensures high efficiency and mechanical reliability for unmanned operations.
• This method is effective for autonomous excavator navigation in challenging environments.