Distributed cooperative Kalman filter constrained by advection-diffusion equation for mobile sensor networks

Area of Science:

• Robotics and Control Systems
• Distributed Systems
• Sensor Networks

Background:

• Mobile sensor networks require accurate state estimation in dynamic environments.
• Spatial-temporal fields, often modeled by advection-diffusion equations, present unique challenges for distributed sensing.

Purpose of the Study:

• To develop a distributed cooperative filtering strategy for state estimation in mobile sensor networks.
• To enable gradient-based formation control for sensor networks operating in spatial-temporal fields.

Main Methods:

• A constrained cooperative Kalman filter was employed for state estimation within a mesh grid of distributed cells.
• The finite volume method was used to discretize the advection-diffusion equation, establishing constraints for the Kalman filters.
• A gradient-based formation control law was developed using the estimated field gradient information.

Main Results:

• The distributed constrained cooperative Kalman filter successfully estimated field values and gradients.
• The gradient-based formation control law effectively adjusted sensor network formation size.
• Convergence analysis confirmed the stability of both the filtering and control strategies.

Conclusions:

• The proposed distributed filtering and control methods are validated for mobile sensor networks.
• This approach offers enhanced state estimation and adaptive formation capabilities in complex spatial-temporal fields.