Event-Based Remote State Estimation for Nonlinear Systems: A Box Particle Filtering Method.

Area of Science:

• Control Systems Engineering
• Signal Processing
• Wireless Communication

Background:

• Remote state estimation for nonlinear/non-Gaussian systems presents challenges due to limited bandwidth in wireless networks.
• Existing methods often rely on Gaussian assumptions and can incur high computational complexity.
• Minimizing data transmission while maintaining estimation accuracy is crucial for practical applications.

Purpose of the Study:

• To develop an event-triggered mechanism for efficient remote state estimation in nonlinear/non-Gaussian systems.
• To design an event-triggered box particle filtering scheme that minimizes mean-squared error.
• To reduce computational complexity and data transmission in wireless networked systems.

Main Methods:

• A novel event-triggering mechanism based on the least-square technique is proposed.
• An event-triggered box particle filtering scheme is designed for minimum mean-squared error estimation.
• Posterior probability density functions are calculated using event-triggered indicator information to manage estimation errors.

Main Results:

• The proposed algorithm effectively reduces data transmissions by employing an event-triggering mechanism.
• It achieves minimum mean-squared error estimation without Gaussian assumptions.
• The method demonstrates reduced computational complexity while ensuring robust estimation performance.

Conclusions:

• The developed event-triggered box particle filtering scheme offers an efficient solution for remote state estimation in nonlinear/non-Gaussian systems.
• The approach successfully addresses bandwidth limitations and computational overhead in wireless networks.
• Simulation results validate the algorithm's effectiveness and performance advantages over existing methods.