Extension of particle filters for time‐varying target presence through split and raw measurements

Target tracking through particle filter (PF) for time‐varying presence of a target is compared for thresholded and non‐thresholded measurements, where in both cases a track produces more than one measurement. To that end, thresholded split measurements PF along with non‐thresholded measurements, sequential importance resampling PF (SIR PF) and auxiliary variable PF (AV PF) are extended to cope with time‐varying target presence. Simulations show superiorities in working through non‐thresholded measurements. Furthermore, they surprisingly demonstrate that non‐thresholded measurements SIR PF leads to less root‐mean‐square position estimation error than non‐thresholded measurements AV PF in case of uncertain target presence and unclear measurements origin besides dynamic model mismatching.