A Novel Particle Filter Based on One-Step Smoothing for Nonlinear Systems with Random One-Step Delay and Missing...

Area of Science:

• Control Systems Engineering
• Signal Processing
• Computational Statistics

Background:

• Networked control systems often face challenges with random one-step delays and missing measurements.
• These issues complicate dynamic state estimation, a critical task in system monitoring and control.
• Existing particle filters, while effective for nonlinear systems, suffer from particle degradation, reducing accuracy.

Purpose of the Study:

• To propose a novel particle filter algorithm designed for nonlinear systems with random one-step delay and missing measurements.
• To mitigate the problem of particle degradation in particle filters.
• To enhance the accuracy and efficiency of state estimation in challenging dynamic environments.

Main Methods:

• A novel particle filter incorporating one-step smoothing is developed.
• The proposed filter iteratively integrates current sensor measurement information into the prior distribution.
• This creates a new importance function to combat particle degeneracy and improve importance sampling efficiency.

Main Results:

• The proposed particle filter effectively limits particle degradation.
• Simulation experiments show improved estimation accuracy compared to the standard bootstrap particle filter.
• The method demonstrates superior performance for nonlinear systems with random one-step delay and missing measurements.

Conclusions:

• The novel particle filter offers a robust solution for state estimation in nonlinear systems with data uncertainties.
• The proposed approach enhances the reliability and accuracy of dynamic state estimation in networked control systems.
• This work contributes to advancing filtering techniques for complex, real-world applications.