Journal of Applied Sciences1812-56541812-5662Asian Network for Scientific Information10.3923/jas.2010.494.499HernaneY. HernaneS. BenyettouM. 62010106Extended Kalman filter is the first algorithm applied to nonlinear state estimation problem and following its limits, other methods based on sampling were developed. We can consider two categories of particle filters: filters which apply a deterministic sampling as the famous unscented Kalman filter and those whose principle is the random sampling as the Particle filter. Furthermore, other approaches that take these two forms of sampling were proposed as Sigma Point Particle filter. The major difficulty of these methods is the computation time which is related to the complexity of sampling. Particle Filter is one among the methods that has attracted particular interest recently; however, PF suffers the problem of degeneration of particles that occurs after re-sampling. We propose to improve PF by the bioinspired algorithm Particle Swarm Optimization as these 2 models have several common. The hybrid method developed in this study is called PFPSO. The PFPSO reduces significantly the degeneracy of the particles; empirical results obtained by applying PFPSO to the problem of estimating the trajectory of a mobile robot illustrate robustness and computational efficiency of our approach.]]>Kalman, R.E., 1960Kennedy, J. and R. Eberhart,1995Julier, S.J. and J.K. Uhlmann,1997Van der Merwe, R., A. Doucet, N. De Freitas and E. Wan,2000Wan, E. and R. Van der Merwe,2001Doucet, A., N. de Freitas and N. Gordon,2001Van der Merwe, R. and E. Wan,2003Clerc, M.,2004Zhang, G., Y. Cheng, F. Yang and Q. Pan,2008