How to arrange the robotic environment? Leveraging experience in both motion planning and environment optimization.

Area of Science:

• Robotics
• Artificial Intelligence
• Optimization Algorithms

Background:

• Robotic system environment design is a high-dimensional optimization problem.
• It involves integrating motion planning (m-DOF) and environment arrangement (free DOF).
• Existing methods struggle with optimal solutions due to complex, high-dimensional search spaces.

Purpose of the Study:

• To propose an experience-based hierarchical-structure optimization algorithm.
• To efficiently solve the combined problem of motion planning and environment arrangement.
• To improve path length and environment arrangement in robotic systems.

Main Methods:

• Developed a hierarchical algorithm with separate layers for environment arrangement (higher) and motion planning (lower).
• Implemented an experience-reuse manner to leverage past results for new problem-solving.
• Proposed the caseGA-ERTC algorithm, utilizing a case-injected genetic algorithm (GA) and experience-driven random tree (ERT).

Main Results:

• Achieved a 12.59% decrease in path lengths for hard problems and 5.05% for easy problems.
• Demonstrated improved environment arrangement and shorter path lengths within the same time constraints.
• Validated effectiveness on three small robot systems compared to state-of-the-art methods.

Conclusions:

• The hierarchical caseGA-ERTC algorithm effectively optimizes robotic system environment design.
• Experience-reuse significantly enhances solution quality and efficiency.
• The proposed method offers a superior approach for complex robotic planning and arrangement tasks.