EFFICIENCY ANALYSIS OF EXTENDED KALMAN FILTERING, UNSCENTED KALMAN FILTERING AND UNSCENTED PARTICLE FILTERING

Abstract

The paper contains algorithms for solving the problem of nonlinear filtering. The nonlinear approximate filters presented are: the extended Kalman filter (EKF), the uncented Kalman filter (UKF) and unscented Particle Filter (UPF). The flow-charts for the listed algorithms are presented and a corresponding software complex based on these algorithms is developed. Using this complex a number of worked-out numerical simulations for both linear and nonlinear models are resulted.

Keywords

• nonlinear filtering
• extended kalman filter
• unscented kalman filter
• unscented particle filter
• sequential monte-carlo estimation