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

Sensor Review ◽  
2021 ◽  
Vol 41 (5) ◽  
pp. 449-457
Author(s):  
Umair Ali ◽  
Wasif Muhammad ◽  
Muhammad Jehanzed Irshad ◽  
Sajjad Manzoor
Keyword(s):  
Neural Network ◽  
Sensor Fusion ◽  
Gaussian Noise ◽  
Computational Cost ◽  
Sensory Information ◽  
Underwater Robot ◽  
Content Type ◽  
Non Gaussian ◽  
Parametric Filter ◽  
Fusion Approach

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.

Robust correntropy ICA based blind channel estimation for MIMO-OFDM systems

2015 ◽  
Vol 34 (3) ◽  
pp. 962-978
Author(s):  
Khaled Abdulaziz Alaghbari ◽  
Lim Heng Siong ◽  
Alan W.C. Tan
Keyword(s):  
Channel Estimation ◽  
Simulation Study ◽  
Gaussian Noise ◽  
Blind Channel Estimation ◽  
Channel Estimator ◽  
Content Type ◽  
Comparison Results ◽  
Blind Channel ◽  
Mimo Ofdm ◽  
Non Gaussian

Purpose – The purpose of this paper is to propose a robust correntropy assisted blind channel estimator for multiple-input multiple-output orthogonal frequency-division multiplexing (MIMO-OFDM) for improved channel gains estimation and channel ordering and sign ambiguities resolution in non-Gaussian noise channel. Design/methodology/approach – The correntropy independent component analysis with L1-norm cost function is used for blind channel estimation. Then a correntropy-based method is formulated to resolve the sign and order ambiguities of the channel estimates. Findings – Simulation study on Gaussian noise scenario shows that the proposed method achieves almost the same performance as the conventional L2-norm based method. However, in non-Gaussian noise scenarios performance of the proposed method significantly outperforms the conventional and other popular estimators in terms of mean square error (MSE). To solve the ordering and sign ambiguities problems, an auto-correntropy-based method is proposed and compared with the extended cross-correlation-based method. Simulation study shows improved performance of the proposed method in terms of MSE. Originality/value – This paper presents for the first time, a correntropy-based blind channel estimator for MIMO-OFDM as well as simulated comparison results with traditional correlation-based methods in non-Gaussian noise environment.

scholarly journals A Novel Approach to 3D-DOA Estimation of Stationary EM Signals Using Convolutional Neural Networks

Sensors ◽  
2020 ◽  
Vol 20 (10) ◽  
pp. 2761
Author(s):  
Dong Chen ◽  
Young Hoon Joo
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Gaussian Noise ◽  
Monitoring Network ◽  
Doa Estimation ◽  
Covariance Matrices ◽  
Noise Environment ◽  
Infinity Norm ◽  
The Neural Network ◽  
Non Gaussian

This paper proposes a novel three-dimensional direction-of-arrival (3D-DOA) estimation method for electromagnetic (EM) signals using convolutional neural networks (CNN) in a Gaussian or non-Gaussian noise environment. First of all, in the presence of Gaussian noise, four output covariance matrices of the uniform triangular array (UTA) are normalized and then fed into four neural networks for 1D-DOA estimation with identical parameters in parallel; then four 1D-DOA estimations of the UTA can be obtained, and finally, the 3D-DOA estimation could be obtained through post-processing. Secondly, in the presence of non-Gaussian noise, the array output covariance matrices are normalized by the infinity-norm and then processed in Gaussian noise environment; the infinity-norm normalization could effectively suppress impulsive outliers and then provide appropriate input features for the neural network. In addition, the outputs of the neural network are controlled by a signal monitoring network to avoid misjudgments. Comprehensive simulations demonstrate that in Gaussian or non-Gaussian noise environment, the proposed method is superior and effective in computation speed and accuracy in 1D-DOA and 3D-DOA estimations, and the signal monitoring network could also effectively control the neural network outputs. Consequently, we can conclude that CNN has better generalization ability in DOA estimation.

Moore-Penrose pseudo-inverse and artificial neural network modeling in performance prediction of switched reluctance machine

2020 ◽  
Vol 39 (6) ◽  
pp. 1411-1430
Author(s):  
Ana Camila Ferreira Mamede ◽  
José Roberto Camacho ◽  
Rui Esteves Araújo ◽  
Igor Santos Peretta
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Computational Cost ◽  
Neural Network Modeling ◽  
Switched Reluctance Machine ◽  
Data Set ◽  
Content Type ◽  
Switched Reluctance ◽  
Reluctance Machine ◽  
Artificial Neural

Purpose The purpose of this paper is to present the Moore-Penrose pseudoinverse (PI) modeling and compare with artificial neural network (ANN) modeling for switched reluctance machine (SRM) performance. Design/methodology/approach In a design of an SRM, there are a number of parameters that are chosen empirically inside a certain interval, therefore, to find an optimal geometry it is necessary to define a good model for SRM. The proposed modeling uses the Moore-Penrose PI for the resolution of linear systems and finite element simulation data. To attest to the quality of PI modeling, a model using ANN is established and the two models are compared with the values determined by simulations of finite elements. Findings The proposed PI model showed better accuracy, generalization capacity and lower computational cost than the ANN model. Originality/value The proposed approach can be applied to any problem as long as experimental/computational results can be obtained and will deliver the best approximation model to the available data set.

Intuitive Intelligent Sensor Fusion With Highly Autonomous Sensors

2001 ◽  
Author(s):  
Jane Xiaojing Yuan ◽  
Fernando Figueroa
Keyword(s):  
Sensor Fusion ◽  
Sensory Information ◽  
Intelligent Sensor ◽  
System A ◽  
Autonomous Sensor ◽  
Fusion Methods ◽  
Qualitative Level ◽  
Fusion Framework ◽  
Different Sources ◽  
Fusion Approach

Abstract The objective of sensor fusion is to synergistically combine different sources of sensory information into one representational format to provide more complete and precise interpretation of the system. A generic sensor fusion framework based on a highly autonomous sensor (HAS) model is presented. The framework provides freedom to choose different data fusion methods and combine them together to achieve better performance. In the context of HAS’s, this paper describes a hierarchical decentralized sensor-fusion approach based on a qualitative theory to interpret measurements, and on qualitative procedures to reason and make decisions based on the measurement interpretations. In this manner, heuristic fusion methods are applied at a high-qualitative level as well as at a numerical level when necessary. This approach implements intuitive (effective) methods to monitor, diagnose, and compensate processes/systems and their sensors.

scholarly journals Efficient steganalysis using convolutional auto encoder network to ensure original image quality

2021 ◽  
Vol 7 ◽  
pp. e356
Author(s):  
Mallikarjuna Reddy Ayaluri ◽  
Sudheer Reddy K. ◽  
Srinivasa Reddy Konda ◽  
Sudharshan Reddy Chidirala
Keyword(s):  
Neural Network ◽  
Gaussian Noise ◽  
Research Method ◽  
Learning Task ◽  
Noise Removal ◽  
Hidden Information ◽  
Computational Overhead ◽  
Research Technique ◽  
Non Gaussian ◽  
Removal Technique

Steganalysis is the process of analyzing and predicting the presence of hidden information in images. Steganalysis would be most useful to predict whether the received images contain useful information. However, it is more difficult to predict the hidden information in images which is computationally difficult. In the existing research method, this is resolved by introducing the deep learning approach which attempts to perform steganalysis tasks in effectively. However, this research method does not concentrate the noises present in the images. It might increase the computational overhead where the error cost adjustment would require more iteration. This is resolved in the proposed research technique by introducing the novel research method called Non-Gaussian Noise Aware Auto Encoder Convolutional Neural Network (NGN-AEDNN). Classification technique provides a more flexible way for steganalysis where the multiple features present in the environment would lead to an inaccurate prediction rate. Here, learning accuracy is improved by introducing noise removal techniques before performing a learning task. Non-Gaussian Noise Removal technique is utilized to remove the noises before learning. Also, Gaussian noise removal is applied at every iteration of the neural network to adjust the error rate without the involvement of noisy features. This proposed work can ensure efficient steganalysis by accurate learning task. Matlab has been employed to implement the method by performing simulations from which it is proved that the proposed research technique NGN-AEDNN can ensure the efficient steganalysis outcome with the reduced computational overhead when compared with the existing methods.

Sign in / Sign up

Export Citation Format

Share Document