Multi-sensor fusion for underwater robot self-localization using PC/BC-DIM neural network

Purpose Self-localization of an underwater robot using global positioning sensor and other radio positioning systems is not possible, as an alternative onboard sensor-based self-location estimation provides another possible solution. However, the dynamic and unstructured nature of the sea environment and highly noise effected sensory information makes the underwater robot self-localization a challenging research topic. The state-of-art multi-sensor fusion algorithms are deficient in dealing of multi-sensor data, e.g. Kalman filter cannot deal with non-Gaussian noise, while parametric filter such as Monte Carlo localization has high computational cost. An optimal fusion policy with low computational cost is an important research question for underwater robot localization. Design/methodology/approach In this paper, the authors proposed a novel predictive coding-biased competition/divisive input modulation (PC/BC-DIM) neural network-based multi-sensor fusion approach, which has the capability to fuse and approximate noisy sensory information in an optimal way. Findings Results of low mean localization error (i.e. 1.2704 m) and computation cost (i.e. 2.2 ms) show that the proposed method performs better than existing previous techniques in such dynamic and unstructured environments. Originality/value To the best of the authors’ knowledge, this work provides a novel multisensory fusion approach to overcome the existing problems of non-Gaussian noise removal, higher self-localization estimation accuracy and reduced computational cost.

Comparison of unit-step responses of parametric filter and fractional-order filter

The paper presents transmission models of a parametric filter with non-periodic variable parameters and a fractional-order filter. The responses of these filters on a unit-step excitation have been examined as well as the dependence of filters time constants on their parameters. The obtained results have been illustrated by examples.

Intelligent FastSLAM: An Intelligent Factorized Solution to Simultaneous Localization and Mapping

FastSLAM is a well-known solution to the simultaneous localization and mapping (SLAM) problem. In FastSLAM, a nonparametric filter is used for the mobile robot pose (position and orientation) estimation, and a parametric filter is used for the feature location's estimation. The performance of the conventional FastSLAM degrades over time due to the particle depletion and unknown statistic noises. In this paper, intelligent FastSLAM (IFastSLAM) is proposed. In this approach, an evolutionary filter (EF) searches stochastically along with the state space for the best robot's pose estimation and an adaptive fuzzy unscented Kalman filter (AFUKF) is used for the feature location's estimation. In AFUKF, a fuzzy inference system (FIS) supervises the performance of the unscented Kalman filter with the aim of reducing the mismatch between the theoretical and actual covariance of the innovation sequences in order to get better consistency. We demonstrate the proposed algorithm with simulations and real-world experiments. The results show that the proposed method is effective, and its performance outperforms conventional FastSLAM.