2014/07/08 by Mamatha Venugopal, Venugopal, Mamatha, R. M. Vasu +3
Computer Science · Earth and Planetary Sciences · Engineering · Environmental Science · #Air Quality Monitoring and Forecasting #Air Quality and Health Impacts #FOS: Computer and information sciences #Methodology (stat.ME) #Structural Health Monitoring Techniques #Target Tracking and Data Fusion in Sensor Networks #Underwater Acoustics Research
paper · pdf · doi:10.48550/arxiv.1407.2192
openalex publication_date 2014/07/08 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Despite the numerous applications that may be expeditiously modelled by\ncounting processes, stochastic filtering strategies involving Poisson-type\nobservations still remain somewhat poorly developed. In this work, we propose a\nMonte Carlo stochastic filter for recursive estimation in the context of\nlinear/nonlinear dynamical systems with Poisson-type measurements. A key aspect\nof the present development is the filter-update scheme, derived from an\nensemble approximation of the time-discretized nonlinear filtering equation,\nmodified to account for Poisson-type measurements. Specifically, the additive\nupdate through a gain-like correction term, empirically approximated from the\ninnovation integral in the filtering equation, eliminates the problem of\nparticle collapse encountered in many conventional particle filters. Through a\nfew numerical demonstrations, the versatility of the proposed filter is brought\nforth, first with application to filtering problems with diffusive or\nPoisson-type measurements and then to an automatic control problem wherein the\nextremization of the associated cost functional is achieved simply by an\nappropriate redefinition of the innovation process.\n