Application of particle filter algorithm based on chaotic sequences and improved t-distribution in UWB indoor...

Area of Science:

• Robotics and Autonomous Systems
• Signal Processing
• Indoor Navigation Technologies

Background:

• Ultra-wideband (UWB) technology is crucial for indoor autonomous navigation and robot path planning.
• Complex environments introduce noise and anomalies in UWB positioning data, degrading accuracy and stability.
• Existing algorithms struggle to mitigate the impact of outliers and noise effectively.

Purpose of the Study:

• To develop an improved particle filter algorithm for robust UWB positioning in challenging indoor settings.
• To enhance the accuracy and stability of UWB positioning by addressing noise and outlier data.
• To provide a computationally efficient solution for real-time UWB localization.

Main Methods:

• Proposed a novel Long-Term Prediction Filter (LTPF) algorithm incorporating chaotic sequences and an improved t-distribution.
• Generated chaotic sequences via a linearly transformed Logistic map to improve particle diversity.
• Constructed an observation likelihood function using standardized residuals and an improved t-distribution to mitigate outlier effects.

Main Results:

• The LTPF algorithm demonstrated significant average positioning accuracy improvements over existing methods (PF, EKPF, KGAPF, SKF-DMM).
• Experimental results confirmed enhanced positioning accuracy and stability in complex indoor environments.
• The LTPF algorithm exhibits computational complexity comparable to the standard Particle Filter (PF).

Conclusions:

• The proposed LTPF algorithm effectively overcomes noise interference and data anomalies in UWB positioning.
• LTPF offers a viable and computationally efficient solution for accurate and stable UWB localization in complex indoor environments.
• This advancement supports more reliable autonomous navigation and robot path planning applications.