Fuzzy adaptive extended Kalman filter for robot 3D pose estimation

Underwater Vehicle Positioning by Correntropy-Based Fuzzy Multi-Sensor Fusion

The ability of the underwater vehicle to determine its precise position is vital to completing a mission successfully. Multi-sensor fusion methods for underwater vehicle positioning are commonly based on Kalman filtering, which requires the knowledge of process and measurement noise covariance. As the underwater conditions are continuously changing, incorrect process and measurement noise covariance affect the accuracy of position estimation and sometimes cause divergence. Furthermore, the underwater multi-path effect and nonlinearity cause outliers that have a significant impact on positional accuracy. These non-Gaussian outliers are difficult to handle with conventional Kalman-based methods and their fuzzy variants. To address these issues, this paper presents a new and improved adaptive multi-sensor fusion method by using information-theoretic, learning-based fuzzy rules for Kalman filter covariance adaptation in the presence of outliers. Two novel metrics are proposed by utilizing correntropy Gaussian and Versoria kernels for matching theoretical and actual covariance. Using correntropy-based metrics and fuzzy logic together makes the algorithm robust against outliers in nonlinear dynamic underwater conditions. The performance of the proposed sensor fusion technique is compared and evaluated using Monte-Carlo simulations, and substantial improvements in underwater position estimation are obtained.

Multi-data sensor fusion framework to detect transparent object for the efficient mobile robot mapping

Purpose

An efficient perception of the complex environment is the foremost requirement in mobile robotics. At present, the utilization of glass as a glass wall and automated transparent door in the modern building has become a highlight feature for interior decoration, which has resulted in the wrong perception of the environment by various range sensors. The perception generated by multi-data sensor fusion (MDSF) of sonar and laser is fairly consistent to detect glass but is still affected by the issues such as sensor inaccuracies, sensor reliability, scan mismatching due to glass, sensor model, probabilistic approaches for sensor fusion, sensor registration, etc. The paper aims to discuss these issues.

Design/methodology/approach

This paper presents a modified framework – Advanced Laser and Sonar Framework (ALSF) – to fuse the sensory information of a laser scanner and sonar to reduce the uncertainty caused by glass in an environment by selecting the optimal range information corresponding to a selected threshold value. In the proposed approach, the conventional sonar sensor model is also modified to reduce the wrong perception in sonar as an outcome of the diverse range measurement. The laser scan matching algorithm is also modified by taking out the small cluster of laser point (w.r.t. range information) to get efficient perception.

Findings

The probability of the occupied cells w.r.t. the modified sonar sensor model becomes consistent corresponding to diverse sonar range measurement. The scan matching technique is also modified to reduce the uncertainty caused by glass and high computational load for the efficient and fast pose estimation of the laser sensor/mobile robot to generate robust mapping. These stated modifications are linked with the proposed ALSF technique to reduce the uncertainty caused by glass, inconsistent probabilities and high load computation during the generation of occupancy grid mapping with MDSF. Various real-world experiments are performed with the implementation of the proposed approach on a mobile robot fitted with laser and sonar, and the obtained results are qualitatively and quantitatively compared with conventional approaches.

Originality/value

The proposed ASIF approach generates efficient perception of the complex environment contains glass and can be implemented for various robotics applications.