Method For Autonomous Underwater Vehicle Research Articles

A Terrain-Following Control Method for Autonomous Underwater Vehicles with Single-Beam Sensor Configuration

This paper investigates the terrain-following problem for an autonomous underwater vehicle (AUV) from control perspectives with full consideration of terrain characteristics. By equipping the AUV with three simple single-beam echo sounders, a set of precise along-track bottom slopes are obtained in a real-time manner, and the occurrence of the lost bottom lock phenomenon caused by a single sonar altimeter is eliminated. A slope-based data processing method is developed, which enables an AUV to characterize seafloor features with complementary sensing modalities to generate proper adaptive height-modified values. In order to keep a fixed height when maneuvering over rugged terrains, a back-stepping depth control is implemented by adjusting horizontal rudder angle, and Lyapunov theory is adopted to analyze the asymptotic stability of the resulting terrain-following control system. At last, simulation results indicate the feasibility and effectiveness of the proposed methodology with a discussion of various sensor configurations.

A High-Precision Positioning Method for Autonomous Underwater Vehicles with Communication Delays

In underwater navigation of autonomous underwater vehicles (AUVs), communication delays frequently occur, leading to a reduction in positioning accuracy. To mitigate this challenge, this work introduces a novel method for relative angle correction, aiming to reconstruct measurement information. Initially, Doppler measurement data are assimilated into the reconstruction of measurement equations to determine the relative angle between the AUV and the observatory. Subsequently, the obtained angle information is integrated into the Extended Kalman Filter (EKF) for the reconstruction of measurement equations. The proposed method effectively reduces positioning errors caused by hydroacoustic communication delays, consequently enhancing AUV positioning accuracy. The efficacy of the proposed method is demonstrated through a simulation study. Simulation results reveal that the incorporation of Doppler angle correction in the reconstructed measurement information method significantly decreases the localization error by approximately 50% compared to EKF and by around 20% compared to the method lacking angle correction.

A chlorophyll a, non‐photochemical fluorescence quenching correction method for autonomous underwater vehicles in shelf sea environments

AbstractAutonomous underwater vehicles provide water column observations of phytoplankton biomass using chlorophyll a (Chl a) fluorometers. However, under high incident light, phytoplankton fluorescence yield decreases in a process known as non‐photochemical quenching, resulting in a reduced Chl a fluorescence signal. Methods have been developed to identify and remove the quenched signal from observations by autonomous underwater vehicles. These existing methods rely on assumptions of the homogeneity of the system, both in terms of time (i.e., between day and night observations) and space (i.e., within the water column or between neighboring profiles). These assumptions are not valid in shallow shelf seas or when sampling across different water masses. Thus, we evaluate six new quenching correction methods based on an existing ocean‐based method, but adapted for a continental shelf sea environment where the water mass changes between night and day observations. We have included two main changes to the existing method. First, we interpolate the unquenched, nighttime signal across the daytime observations and use this as a reference for correcting the quenched, daytime signal. Second, we explore the inclusion of a fluorescence quenching depth limit. By interpolating nighttime observations across daytime periods, the diel changes in non‐photochemical quenching were separated from the phytoplankton population changes. The proposed methods all show improved performance compared to existing approach. The methods presented here, and the approach used to evaluate them, are applicable in other shelf sea environments, enabling studies using autonomous Chl a fluorescence data across shelf sea ecosystems.

Data-Driven Fault Detection of AUV Rudder System: A Mixture Model Approach

Based on data-driven and mixed models, this study proposes a fault detection method for autonomous underwater vehicle (AUV) rudder systems. The proposed method can effectively detect faults in the absence of angle feedback from the rudder. Considering the parameter uncertainty of the AUV motion model resulting from the dynamics analysis method, we present a parameter identification method based on the recurrent neural network (RNN). Prior to identification, singular value decomposition (SVD) was chosen to denoise the original sensor data as the data pretreatment step. The proposed method provides more accurate predictions than recursive least squares (RLSs) and a single RNN. In order to reduce the influence of sensor parameter errors and prediction model errors, the adaptive threshold is mentioned as a method for analyzing prediction errors. In the meantime, the results of the threshold analysis were combined with the qualitative force analysis to determine the rudder system’s fault diagnosis and location. Experiments conducted at sea demonstrate the feasibility and effectiveness of the proposed method.

Lagrange tracking-based long-term drift trajectory prediction method for Autonomous Underwater Vehicle.

Autonomous Underwater Vehicle (AUV) works autonomously in complex marine environments. After a severe accident, an AUV will lose its power and rely on its small buoyancy to ascend at a slow speed. If the reserved buoyancy is insufficient, when reaching the thermocline, the buoyancy will rapidly decrease to zero. Consequently, the AUV will experience prolonged lateral drift within the thermocline. This study focuses on developing a prediction method for the drift trajectory of an AUV after a long-term power loss accident. The aim is to forecast the potential resurfacing location, providing technical support for surface search and salvage operations of the disabled AUV. To the best of our knowledge, currently, there is no mature and effective method for predicting long-term AUV underwater drift trajectories. In response to this issue, based on real AUV catastrophes, this paper studies the prediction of long-term AUV underwater drift trajectories in the cases of power loss. We propose a three-dimensional trajectory prediction method based on the Lagrange tracking approach. This method takes the AUV's longitudinal velocity, the time taken to reach different depths, and ocean current data at various depths into account. The reason for the AUV's failure to ascend to sea surface lies that the remaining buoyancy is too small to overcome the thermocline. As a result, AUV drifts long time within the thermocline. To address this issue, a method for estimating thermocline currents is proposed, which can be used to predict the lateral drift trajectory of the AUV within the thermocline. Simulation is conducted to compare the results obtained by the proposed method and that in a real accident. The results demonstrate that the proposed approach exhibits small directional and positional errors. This validates the effectiveness of the proposed method.

Adaptive non-singular integral terminal sliding mode-based fault tolerant control for autonomous underwater vehicles

This article proposes an adaptive non-singular integral terminal sliding mode-based fault tolerant control (ANITSM-FTC) method for autonomous underwater vehicles (AUVs) exposed to uncertainty of dynamics and potential unknown thruster failures. The developed control scheme shows unique superiority in convergence rate over the existing adaptive proportional–integral sliding mode-based fault tolerant control (APISM-FTC), and overcomes the singularity problem effectively compared to traditional sliding mode-based scheme. During the design of the ANITSM-FTC controller, radial basis function neural network (RBFNN) and adaptive technique are adopted to estimate the dynamics uncertainty. Rigorous theoretical analysis and rigorous mathematical proof based on the trajectory tracking problems of AUVs are implemented, to ensure all tracking errors asymptotically converge to zero. To more fully demonstrate the viability of the proposed ANITSM-FTC method, the simulation experiments are carried out with three different thruster failure conditions by a fair comparison of APISM-FTC. The proposed scheme inherits the advantages of dealing with the problem of possible thruster failures as well as overcoming the drawbacks of other methods.

Hybrid model navigation method for autonomous underwater vehicle

Accurate navigation and localization are essential for Autonomous Underwater Vehicles (AUVs). However, the unknown modeling errors and nonlinear errors will affect the AUV positioning accuracy. Meanwhile, the marine environment changes may not be accurately sensed by AUV. Therefore, this paper proposes a Hybrid Model (HM) navigation methodology for AUV to reduce the impact of unknown errors and better predict the future states simultaneously. Firstly, an error correction sub-model based on Sequence to Sequence (Seq2Seq) predicts AUV pseudo displacements. The pseudo displacements are augmented to the observation vector to correct the unknown errors in the state estimation. Secondly, a state regression sub-model based on Gaussian Process Regression (GPR) is utilized to capture the motion trends from the historical data and regress the state variation, which is used to resist the unknown disturbances in the changeable marine environment. The sub-models work parallel with the master model, benefiting from the Interacting Multiple Model (IMM). We compare the performance of the proposed HM to Extended Kalman Filter (EKF), Unscented Kalman Filter (UKF), and IMM-EKF by using the actual experimental data of Sailfish 210 AUV. The experimental results show that the proposed HM algorithm achieves superior navigation accuracy and good fault tolerance capability.

Occupancy Grid-Based AUV SLAM Method with Forward-Looking Sonar

Simultaneous localization and mapping (SLAM) is an active localization method for Autonomous Underwater Vehicle (AUV), and it can mainly be used in unknown and complex areas such as coastal water, harbors, and wharfs. This paper presents a practical occupancy grid-based method based on forward-looking sonar for AUV. The algorithm uses an extended Kalman filter (EKF) to estimate the AUV motion states. First, the SLAM method fuses the data coming from the navigation sensors to predict the motion states. Subsequently, a novel particle swarm optimization genetic algorithm (PSO-GA) scan matching method is employed for matching the sonar scan data and grid map, and the matching pose would be used to correct the prediction states. Lastly, the estimated motion states and sonar scan data would be used to update the grid map. The experimental results based on the field data have validated that the proposed SLAM algorithm is adaptable to underwater conditions, and accurate enough to use for ocean engineering practical applications.

Smart vector-inspired optical vision guiding method for autonomous underwater vehicle docking and formation.

A smart vector-inspired optical vision guiding (VIOVG) method for autonomous underwater vehicle (AUV) docking and formation is proposed. Unlike traditional optical guiding methods based on LED arrays, the method is inspired by a vector, and uses four laser diodes to form a wing-light pattern to realize smart optical guiding for AUVs. Due to the light scattering effect from water, the four laser diodes can emit four bright and slim laser beams in water, which can be captured by underwater cameras. The intersections of the laser beams form wing-lights as markers for AUV distance and pose estimation. The wing-light pattern is easily tuned and extended by changing the propagation direction and the power of the laser beams beyond the limitation of the physical size of the AUV or docking station. The simulations and experiments show that the proposed method can achieve high precision positioning. A relative distance error of 3.35% is achieved in a positioning experiment at ∼10 m. It has great potential for AUV docking and formation, especially for small AUVs.

Numerical Simulation of Time-Optimal Path Planning for Autonomous Underwater Vehicles Using a Markov Decision Process Method

Many path planning algorithms developed for land or air based autonomous vehicles no longer apply under the water. A time-optimal path planning method for autonomous underwater vehicles (AUVs), based on a Markov decision process (MDP) algorithm, is proposed for the marine environment. Its performance is examined for different oceanic conditions, including complex coastal bathymetry and time-varying ocean currents, revealing advantages compared to the A* algorithm, a traditional path planning method. The ocean current is predicted using a regional ocean model and then provided to the MDP algorithm as a priori. A computation-efficient and feature-resolved spatial resolution are determined through a series of sensitivity experiments. The simulations demonstrate the importance to incorporate ocean currents in the path planning of AUVs in the real ocean. The MDP algorithm remains robust even if the ocean current is complex.

Event-Based Circular Detection for AUV Docking Based on Spiking Neural Network.

In this paper, a circular objects detection method for Autonomous Underwater Vehicle (AUV) docking is proposed, based on the Dynamic Vision Sensor (DVS) and the Spiking Neural Network (SNN) framework. In contrast to the related work, the proposed method not only avoids motion blur caused by frame-based recognition during docking procedure but also reduces data redundancy with limited on-chip resources. First, four coplanar and rectangular constrained circular light sources are constructed as the docking landmark. By combining asynchronous Hough circle transform with the SNN model, the coordinates of landmarks in the image are detected. Second, a Perspective-4-Point (P4P) algorithm is utilized to calculate the relative pose between AUV and the landmark. In addition, a spatiotemporal filter is also used to eliminate noises generated by the background. Finally, experimental results are demonstrated from both software simulation and experimental pool, respectively, to verify the proposed method. It is concluded that the proposed method achieves better performance in accuracy and efficiency in underwater docking scenarios.

Carbon dioxide treatment method for autonomous underwater vehicles powered by direct methanol fuel cells: A multi-criteria decision analysis approach

Autonomous Underwater Vehicles are very valuable tools in a great variety of marine related sectors that demand new capacities and improvements. These improvements go through increasing their endurance and, in this sense, the use of Direct Methanol Fuel Cells can be very interesting with an adequate CO2 Treatment Method. This work investigates four CO2 Treatment Methods for their application onboard Autonomous Underwater Vehicles powered by a Direct Methanol Fuel Cell-based power plant. The evaluation was carried out considering different working conditions defined by the fuel cell power, navigation depth and endurance. To rank the treatment methods, three Multi Criteria Decision Methods and eight selection criteria were used. The results show that direct disposal of CO2 ejecting it outside the Autonomous Underwater Vehicle is the most adequate method for navigation depths up to 50 m. For deeper waters CO2 storage embedded in an adsorbent material and CO2 stored as a pressurized gas seem to be the best choices. In all cases, the results present a good stability against changes in the criteria weights vector.

Underwater-Sonar-Image-Based 3D Point Cloud Reconstruction for High Data Utilization and Object Classification Using a Neural Network

This paper proposes a sonar-based underwater object classification method for autonomous underwater vehicles (AUVs) by reconstructing an object’s three-dimensional (3D) geometry. The point cloud of underwater objects can be generated from sonar images captured while the AUV passes over the object. Then, a neural network can predict the class given the generated point cloud. By reconstructing the 3D shape of the object, the proposed method can classify the object accurately through a straightforward training process. We verified the proposed method by performing simulations and field experiments.

Integral vector field control for three-dimensional path following of autonomous underwater vehicle

A new control method for autonomous underwater vehicle (AUV) is developed to achieve three-dimensional (3D) path following under ocean current disturbances when AUV lacks lateral and vertical driving forces. To make the control system more convenient for practical application and reduce the energy consumption of AUV, the controller is designed with relative velocities. To avoid the singularity problem of curve path following, the Serret–Frenet frame is introduced as virtual target and the error model is defined. The adaptive law of the virtual target and an integral vector field (IVF) guidance law are designed in the kinematic controller. The back-stepping method and adaptive dynamical sliding mode control (BADSMC) technology are applied in the design of dynamic controller. The Lyapunov theory and nonlinear cascade system theory are applied to prove the closed-loop stability. Finally, the performances of the IVF + BADSMC and ILOS + PID algorithm are compared through simulation. Simulation results demonstrate that the proposed controller can realize the path following of AUV under disturbances of ocean current and can improve the following quality.

Multisensor online 3D view planning for autonomous underwater exploration

AbstractThis study presents a novel octree‐based three‐dimensional (3D) exploration and coverage method for autonomous underwater vehicles (AUVs). Robotic exploration can be defined as the task of obtaining a full map of an unknown environment with a robotic system, achieving full coverage of the area of interest with data from a particular sensor or set of sensors. While most robotic exploration algorithms consider only occupancy data, typically acquired by a range sensor, our approach also takes into account optical coverage, so the environment is discovered with occupancy and optical data of all discovered surfaces in a single exploration mission. In the context of underwater robotics, this capability is of particular interest, since it allows one to obtain better data while reducing operational costs and time. This study expands our previous study in 3D underwater exploration, which was demonstrated through simulation, presenting improvements in the view planning (VP) algorithm and field validation. Our proposal combines VP with frontier‐based (FB) methods, and remains light on computations even for 3D environments thanks to the use of the octree data structure. Finally, this study also presents extensive field evaluation and validation using the Girona 500 AUV. In this regard, the algorithm has been tested in different scenarios, such as a harbor structure, a breakwater structure, and an underwater boulder.

Tracking a Sea Turtle by an AUV with a Multibeam Imaging Sonar: Toward Robotic Observation of Marine Life

This paper proposes a method for autonomous underwater vehicles to chase sea turtles without attaching any tag to them, toward efficient and long-term observation of marine life. The method uses a multibeam imaging sonar as the main sensor to detect sea turtles. The method utilizes convolutional neural network (CNN) for detecting a sea turtle in sonar imagery. Surge and yaw movements of the vehicle are controlled to maintain the relative distance and direction to the detected target. The proposed method was implemented in the AUV HATTORI. The AUV succeeded in tracking a sea turtle in natural condition for 270 seconds in shallow sea.

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

This paper presents a novel navigation method for Autonomous Underwater Vehicle (AUV). In respect of improving the precision of navigation, recent research mainly focused on how to reduce the system model error. However, existing optimization methods are less efficient and effective in decreasing interference of sensor deviation. Motivated by the excellent performance of deep-learning, a Hybrid Recurrent Neural Networks (Hybrid RNNs) framework is proposed to estimate the AUV position. Firstly, since the different sensors have different data frequency, this method employs unidirectional and bi-directional long short-term memory (LSTM) with multiple memory units to handle raw sensor values in a single calculation cycle. Subsequently, using the outputs of LSTMs and the time interval of the cycle above, the fully connected layers could obtain the displacements of AUV. Eventually, to verify the effectiveness of the proposed navigation algorithm, a series of evaluations have been carried out, which are based on a public dataset and real experimental data of our AUV. The evaluation results have been validated that the proposed method can reduce the interference of sensor deviation, and has better accuracy as well as fault tolerance for navigation. Meanwhile, it could also satisfy the real-time requirement.

Constrained Control of Autonomous Underwater Vehicles Based on Command Optimization and Disturbance Estimation

In this paper, a method is presented for antidisturbance constrained control of autonomous underwater vehicles subject to uncertainties and constraints. The uncertainties stem from uncertain hydrodynamic parameters, modeling errors, and unknown forces due to the ocean currents in an underwater environment. An antidisturbance constrained controller is developed by designing a command governor and a disturbance observer. Specifically, the disturbance observer is developed to estimate the lumped disturbance composed of parametric model uncertainties, modeling errors, and unknown environmental forces. The command governor is designed for optimizing command signals in the receding horizon within the state and input constraints. The command governor is formulated as a quadratically constrained quadratic programming problem. To facilitate online implementations, a neurodynamic optimization method based on a one-layer recurrent neural network is employed for solving the quadratic optimization problem subject to inequality constraints with finite-time convergence. The efficacy of the proposed antidisturbance constrained control method for autonomous underwater vehicles is substantiated via simulations and comparisons.

Distributed sensing for fluid disturbance compensation and motion control of intelligent robots

A control methodology for aerial or aquatic vehicles is presented that leverages intelligent distributed sensing inspired by the lateral line found in fish to directly measure the fluid forces acting on the vehicle. As a result, the complex robot control problem is effectively simplified to that of a rigid body in a vacuum. Furthermore, by sensing these forces, they can be compensated for immediately, rather than after they have displaced the vehicle. We have created a sensory shell around a prototype autonomous underwater vehicle, derived algorithms to remove static pressure and calculate total force from the discrete measurements using a fitting technique that filters sensor error, and validated the control methodology on a vehicle in the presence of multiple fluid disturbances. This sensing control scheme reduces position tracking errors by as much as 72% compared to a standard position error feedback controller. Accurate manoeuvring of autonomous aerial and aquatic robots requires detailed knowledge of the fluid forces, which can be challenging especially in turbulent water or air. A control method for autonomous underwater vehicles (AUVs) uses intelligent distributed sensing inspired by fish ‘lateral line’ sensing. This is used by many species of fish to feel the flow around them and respond instantly, before they are displaced by disturbances. An AUV designed with such a sensory shell similarly compensates for disturbances and has improved position tracking.

Autonomous Tracking of Sea Turtles based on Multibeam Imaging Sonar: Toward Robotic Observation of Marine Life

This paper proposes a method for autonomous underwater vehicles to track sea turtles without attaching any tag to them, toward efficient and long-term observation of marine life. The method uses a multibeam imaging sonar as the main sensor to detect sea turtles. The method utilizes convolutional neural network (CNN) for detecting a sea turtle in sonar imagery. Surge and yaw movements of the vehicle are controlled to maintain the relative distance and direction to the detected target. The proposed method was implemented in the AUV HATTORI. The AUV succeeded in tracking a sea turtle in natural condition for 270 seconds in shallow sea.