An Improved Particle Filtering Algorithm Using Different Correlation Coefficients for Nonlinear System State...

Area of Science:

• Signal Processing
• Statistical Inference
• Computational Mathematics

Background:

• Particle filtering (PF) is widely used in nonlinear and non-Gaussian scenarios.
• Particle degeneracy is a significant limitation in traditional PF algorithms.
• Existing resampling methods aim to mitigate degeneracy but have limitations.

Purpose of the Study:

• To propose an improved particle filtering algorithm to overcome particle degeneracy.
• To enhance the accuracy and robustness of particle filtering in challenging noise environments.
• To evaluate the performance of the proposed algorithm against established PF techniques.

Main Methods:

• Developed an improved particle filtering algorithm incorporating rank correlation coefficients.
• Implemented and simulated the algorithm using Matlab.
• Compared the proposed algorithm with Sequential Importance Resampling (SIR), Gaussian Sum Particle Filter (GSPF), and Gaussian Mixture Sigma-Point Particle Filters (GM-SPPF).

Main Results:

• The proposed improved PF algorithm effectively addresses particle degeneracy.
• The algorithm demonstrates significantly better accuracy in Gaussian mixture noise.
• Performance evaluation confirmed superior accuracy compared to SIR, GSPF, and GM-SPPF.

Conclusions:

• The novel PF algorithm integrating rank correlation coefficients offers a robust solution for degeneracy.
• This method provides enhanced accuracy for particle filtering in non-Gaussian noise conditions.
• The findings suggest a valuable advancement for applications requiring precise state estimation.