Simulation of a Radar-Based 2D Object Tracking and Velocity Estimator in the X-Y Plane

Abstract

This study tackles the significant challenge of augmenting accuracy in the two-dimensional (2D) tracking of mobile objects, particularly in environments plagued by high noise and data uncertainty. Traditional Kalman filter-based methods often struggle under these conditions, failing to deliver the needed precision. Consequently, the primary objective of this research is to devise, implement, and
validate a refined linear Kalman filter approach. This approach aims to significantly diminish estimation errors in tracking object positions
and velocities, addressing the identified limitations of existing methodologies. The approach involves an innovative adaptation of the linear Kalman filter technique, rigorously evaluated through detailed simulations and analysis to ascertain its effectiveness in enhancing trackingaccuracy under challenging conditions. The results show that a linear Kalman Filter can be used to accurately estimate the position of mobile objects, with root mean square errors of less than 0.1%. We demonstrate marked improvements in accuracy and reliability for 2D object tracking, showcasing the method's potential applicability in real-world scenarios such as autonomous navigation, wildlife and automobile tracking systems. While recognizing the persistent challenges posed by noise variations, this research paves the way for future exploration into nonlinear applications and update processes. Such investigations are anticipated to further refine object tracking methodologies, building on the foundational work presented here.