EXPLORATION AND DATA REFINEMENT VIA MULTIPLE MOBILE SENSORS BASED ON GAUSSIAN PROCESSES

Area of Science:

• Robotics
• Artificial Intelligence
• Sensor Networks

Background:

• Mobile sensor networks are crucial for data collection in unknown environments.
• Balancing exploration of new areas and refinement of existing data presents a significant challenge.
• Local sensor data limits the ability to determine globally optimal next steps.

Purpose of the Study:

• To develop a framework for optimizing mobile sensor trajectories.
• To address the conflicting goals of broad field exploration and detailed data refinement.
• To enable intelligent decision-making for sensor movement in unknown environments.

Main Methods:

• Utilized Gaussian process regression for data analysis and prediction.
• Developed a framework to integrate exploration and data refinement objectives.
• Implemented a decision-making process for mobile sensor trajectory planning.

Main Results:

• The proposed framework effectively balances exploration and refinement goals.
• Gaussian process regression provides a viable method for trajectory optimization.
• Demonstrated reasonable decision-making for sensor movement based on local information.

Conclusions:

• The framework offers a practical solution for mobile sensor configuration in unknown fields.
• Gaussian process regression is a powerful tool for multi-objective sensor navigation.
• Future work can extend this approach to more complex scenarios and sensor types.