Efficient Autonomous Exploration and Mapping in Unknown Environments.

Area of Science:

• Robotics
• Artificial Intelligence
• Autonomous Systems

Background:

• Autonomous exploration and mapping are crucial for robots in unknown environments.
• Current methods overlook regional legacy issues, reducing long-term exploration efficiency.

Purpose of the Study:

• To propose a novel strategy addressing regional legacy issues in autonomous exploration.
• To enhance the efficiency and adaptability of robot exploration in unknown environments.

Main Methods:

• Introduced the Local-and-Global Strategy (LAGS) algorithm.
• Integrated Gaussian process regression (GPR), Bayesian optimization (BO), and deep reinforcement learning (DRL).
• Combined local exploration with global perception strategies.

Main Results:

• The proposed method significantly improves exploration efficiency.
• Shorter paths and higher adaptability were demonstrated across diverse maps.
• Ensured robot safety during exploration.

Conclusions:

• The LAGS algorithm effectively solves regional legacy issues in autonomous exploration.
• The integrated GPR, BO, and DRL models enhance exploration performance and safety.
• The method shows strong adaptability for various unknown environments.