A Soar-Based Space Exploration Algorithm for Mobile Robots

Area of Science:

• Robotics
• Artificial Intelligence
• Cognitive Science

Background:

• Mobile robot space exploration is crucial but challenged by dynamic environments.
• Existing methods like frontier exploration, heuristic algorithms, and deep reinforcement learning face limitations in real-time performance.
• The cognitive architecture of Soar offers a potential framework for advanced decision-making in robots.

Purpose of the Study:

• To model mobile robot space exploration using the Soar cognitive architecture.
• To develop and evaluate novel heuristic algorithms (HAs) for faster space exploration.
• To enhance the speed and efficiency of mobile robots in dynamic environments.

Main Methods:

• Modeling the space exploration problem within the Soar cognitive architecture.
• Proposing three new heuristic algorithms (HAs) based on the Soar model.
• Conducting experimental validation in the Easter environment.

Main Results:

• The proposed heuristic algorithms (HAs) demonstrated a significant improvement in exploration speed.
• The Easter robot's exploration speed was enhanced by at least 2.04 times compared to the original algorithm.
• The effectiveness of the HAs in accelerating robot space exploration was verified.

Conclusions:

• The Soar cognitive architecture provides a robust model for mobile robot space exploration.
• The developed heuristic algorithms (HAs) effectively increase exploration speed in dynamic environments.
• This strategy offers a promising solution for time-critical space exploration tasks for mobile robots.