A robust cooperative localization algorithm based on covariance intersection method for multi-robot systems.

Area of Science:

• Robotics
• Artificial Intelligence
• Computer Science

Background:

• Cooperative localization is crucial for multi-robot systems, particularly when reducing base station communication load is necessary.
• Existing methods often face challenges with measurement uncertainty and linearization errors.

Purpose of the Study:

• To propose a novel cooperative localization algorithm for multi-robot systems.
• To achieve high-accuracy localization by utilizing relative measurements among robots.
• To address uncertainty and avoid linearization errors inherent in traditional methods.

Main Methods:

• A particle-based approximation method is employed to handle uncertainty in robot positions and mitigate linearization errors.
• Covariance intersection is used to fuse preliminary estimations from different robots, ensuring a minimum upper bound for fused covariance.
• Kullback-Leibler divergence is utilized to detect and reject abnormal measurements, preventing their negative impact on localization accuracy.

Main Results:

• The proposed algorithm demonstrates higher localization accuracy compared to three other algorithms in simulations.
• The method effectively handles abnormal measurements, improving the robustness of the cooperative localization system.
• Simulations validate the algorithm's performance in complex multi-robot scenarios.

Conclusions:

• The novel cooperative localization algorithm offers improved accuracy and robustness for multi-robot systems.
• The integration of particle-based approximation and Kullback-Leibler divergence provides an effective solution for handling measurement uncertainties and anomalies.
• This approach is particularly beneficial for applications requiring reduced communication load and high-precision positioning.