Robust particle filter formulations with application to terrain‐aided navigation

Abstract

Summary The work described in this paper is motivated by the need to develop efficient and robust estimation filters with application to terrain‐aided navigation of underwater robotic vehicles. One of the main problems addressed is the development of navigation particle filters that can deal with the scarcity of landmarks and the terrain ambiguity that characterize vast regions of the ocean floor. As a contribution to solve this problem, the paper proposes three novel particle filter algorithms and assesses their estimation efficiency and robustness to non‐informative measurements using two well‐known benchmarking tests. The performance of the new filters in these tests demonstrates their potential to solve a class of nonlinear problems that include, but are not limited to, the type of underwater navigation problem that motivated the present work. Our study concludes by examining the performance of the filters in terms of determining the position and velocity of an autonomous underwater vehicle in the presence of unknown ocean currents. When applied to terrain‐aided navigation, the novel particle filter formulations formulations mitigate filter divergence issues frequently caused by terrain symmetries and are more robust than other well‐known versions when used in scenarios with poor terrain information. The theoretical developments presented and the results obtained in simulations are validated using real data acquired during tests with an autonomous marine robot. Copyright © 2016 John Wiley & Sons, Ltd.

A data‐driven particle filter for terrain based navigation of sensor‐limited autonomous underwater vehicles

“…Simulation results demonstrated the superiority of the PPF in terms of the asymptotic convergence of the filters. Later, a suitable compact support for the uniform distributions was derived, using the Fisher information matrix of the terrain . In a similar approach, in a PF is also used, with subsets of the particles being weighted with different weighting functions.…”

“…The choice of the hyperparameters can play a relevant role in the prediction process. In the case of a covariance function given by (12), = {l, f , n }. While the values for each element of could be made empirically, given some knowledge of the system, a more dynamic strategy could be adopted by maximizing the log likelihood of the training outputs given the inputs.…”

Design and Implementation of the RSIL to LLVM IR Translator for Verification of the Intermediate Code on IoT Virtual Machine

“…With the rapid advancement in location technologies and tracking devices, and the demand for flawless solutions to overcome the problems associated with current mobile location based techniques, there is widespread interest in mobile IPS systems (IPS) [1][2][3][4][5][6][7][8][9][10][11][12][13]. One of the major components of mobile IPS is inertial-based positioning, which facilitates the tracing of individuals (or mobile nodes) within corridors or other enclosed structures by using inertial sensor.…”

“…This phenomenon will cause computation burden to the overall systems [24]. Two (2) reason behind this issue, whether it caused by the of small noise measurement in accurate sensor or small number of particle [25]. In order to counter the phenomenon, the special strategies resampling algorithm can be used.…”

Adaptive memory-based single distribution resampling for particle filter

“…In such models, the noises can be non-Gaussian. Several fields have adopted this methodology including: finance [5][6][7][8], wireless communications [9][10][11][12], geophysical systems [13][14][15][16][17], navigation and tracking [18][19][20], control [21][22][23][24][25], and robotics [26][27][28][29][30][31]. Generally, this methodology can approximate state density p(x k ) using a range of random particles that have related nonnegative weights:…”