Advanced point-mass method for nonlinear state estimation

Computing arrival cost parameters in moving horizon estimation using sampling based filters

“…(12) and (14) and obtain the parameters of arrival cost for MHE. The tra ditional MHE initialization by EKF is also included for comparison with the deterministic sampling based unscented Kalman filter, the random sampling based particle filter and the aggregate Markov chain based cell filter.…”

“…They include the deterministic sampling based unscented Kalman filter (UKF), the random sampling based class of nonlinear filters called particle filter (PF) and the aggregate Markov chain based cell filter (CF). The choice of these three methods is motivated by their rela tively small online computational demand compared to other non traditional filters such as the grid based approaches [9,[12][13][14].…”

“…(12) and (14). No assumptions are made about the shape of the conditional pdf, which is approximately represented by a fi nite set of random samples and associated weights.…”

SONAR Systems and Underwater Signal Processing: Classic and Modern Approaches

“…At each time update, new particles are generated via sampling from an importance function, and the weight associated with each particle is updated. With fast development of powerful computational systems, the numerical approach to solve the Bayesian inference equation that was first introduced in (Bucy 1971) has renewed its attraction through the point mass approach (Simandl 2006). In the point mass approach, a regular grid discretizes the system state, and the Bayesian filter is evaluated numerically on the discrete points.…”

The marginalized particle filter – analysis, applications and generalizations

“…This system was used in [38], where it illustrated grid-based (point-mass) filters. Obviously, the states can be estimated by applying the standard particle filter to the entire state vector.…”