End-to-end navigation for Autonomous Underwater Vehicle with Hybrid Recurrent Neural Networks

Mu, Xiaokai; Zhang, Xin; Yan, Shuicheng; Shen, Yue; Feng, Chen

doi:10.1016/j.oceaneng.2019.106602

Cited by 37 publications

(7 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In particular, when monitoring attributes and hydrological factors are combined as system inputs, the prediction result reaches the best. To verify the advantages of the constructed model, we compare it with multiple traditional regression models, including Linear Regression (LR), Robust Linear Regression (RLR), Interaction Linear Regression (IR), Pure Quadratic Regression (PQR) and Fine Tree Regression (FTR) (Yang, 2018;Goebel and Plötz, 2019;Acharya et al, 2019), BPNN (Back Propagation Neural Network) and DBPNN (BPNN with two hidden layers), and some common RNN networks, including general RNN, Bidirectional RNN (BRNN) and Deep RNN (DRNN) (Cui et al, 2018;Mu et al, 2019). In the comparison experiments, we select 10 feature variables as inputs of models and use the same training (12416 data records) and testing data (3105 data records).…”

Section: Case Studymentioning

confidence: 99%

See 1 more Smart Citation

Prediction and optimisation of fuel consumption for inland ships considering real-time status and environmental factors

et al. 2021

View full text Add to dashboard Cite

Section: Case Studymentioning

confidence: 99%

“…It has been widely used in data processing and modelling in water transport Lin et al (2019). proposed a RNN with convolution for online obstacle avoidance in unmanned underwater vehicles Mu et al (2019). proposed Hybrid RNNs that use unidirectional and bi-directional LSTMs to handle different sensor data.…”

mentioning

confidence: 99%

Prediction and optimisation of fuel consumption for inland ships considering real-time status and environmental factors

et al. 2021

View full text Add to dashboard Cite

“…However, they did not mention the training time requirement, processing load, and the total number of neurons used in the network. Another current study proposed an end-to-end navigation solution based on deep hybrid recurrent neural networks and used raw sensors data directly to estimate the location underwater vehicle [24]. An investigation that was conducted recently took advantage of Reinforcement Learning (RL) and incorporated the deep deterministic policy gradient for tuning the process noise covariance matrix online from low-cost navigational sensors [25].…”

Section: Review Of Previous Workmentioning

confidence: 99%

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

Shaukat

Moinuddin

Otero

2021

Sensors

View full text Add to dashboard Cite

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.

show abstract

“…Therefore, we employed a hybrid recurrent neural networks (RNNs) framework to realize AUV position estimation. Since the training process uses the GPS movement as the label, and the raw data of the sensors as the input, the trained framework could include the interference from yaw error [ 32 ]. Figure 3 presents the structure of hybrid RNNs.…”

Section: Adaptive Navigation Algorithm With Deep Learningmentioning

confidence: 99%

“…Since the neural network method could approach the optimal solution, this method could effectively improve the navigation accuracy when the sensor has a large deviation. However, according to our research, the positioning accuracy of the neural network method is unsatisfactory compared to that of the traditional navigation method when the sensor accuracy is high [ 32 ]. Additionally, the navigation information frequency of the deep learning method is low since the calculation cycle of the deep learning method needs data from a duration time, which would cause this method to be insufficient in some missions.…”

Section: Introductionmentioning

confidence: 99%

Adaptive Navigation Algorithm with Deep Learning for Autonomous Underwater Vehicle

2021

Sensors

Self Cite

View full text Add to dashboard Cite

Precise navigation is essential for autonomous underwater vehicles (AUVs). The measurement deviation of the navigation sensors, especially the microelectromechanical systems (MEMS) sensors, is a crucial factor that affects the localization accuracy. Deep learning is a novel method to solve this problem. However, the calculation cycle and robustness of the deep learning method may be insufficient in practical application. This paper proposes an adaptive navigation algorithm with deep learning to address these questions and realize accurate navigation. Firstly, this algorithm uses deep learning to generate low-frequency position information to correct the error accumulation of the navigation system. Secondly, the χ2 rule is selected to judge if the Doppler velocity log (DVL) measurement fails, which could avoid interference from DVL outliers. Thirdly, the adaptive filter, based on the variational Bayesian (VB) method, is employed to estimate the navigation information simultaneous with the measurement covariance, improving navigation accuracy even more. The experimental results, based on AUV field data, show that the proposed algorithm could realize robust navigation performance and significantly improve position accuracy.

show abstract

End-to-end navigation for Autonomous Underwater Vehicle with Hybrid Recurrent Neural Networks

Cited by 37 publications

References 15 publications

Prediction and optimisation of fuel consumption for inland ships considering real-time status and environmental factors

Prediction and optimisation of fuel consumption for inland ships considering real-time status and environmental factors

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

Adaptive Navigation Algorithm with Deep Learning for Autonomous Underwater Vehicle

Contact Info

Product

Resources

About