Underwater Vehicle Research Articles

Deep‐Learning‐Assisted Triboelectric Whisker Sensor Array for Real‐Time Motion Sensing of Unmanned Underwater Vehicle

AbstractAquatic animals can perceive their surrounding flow fields through highly evolved sensory systems. For instance, a seal whisker array understands the hydrodynamic field that allows seals to forage and navigate in dark environments. In this work, a deep learning‐assisted underwater triboelectric whisker sensor array (TWSA) is designed for the 3D motion estimation and near‐field perception of unmanned underwater vehicles. Each sensor comprises a high aspect ratio elliptical whisker shaft, four sensing units at the root of the elliptical whisker shaft, and a flexible corrugated joint simulating the skin on the cheek surface of aquatic animals. The TWSA effectively identifies flow velocity and direction in the 3D underwater environments and exhibits a rapid response time of 19 ms, a high sensitivity of 0.2V/ms−1, and a signal‐to‐noise ratio of 58 dB. The device also locks onto the frequency of the upstream wake vortex, achieving a minimal detection accuracy of 81.2%. Moreover, when integrated with an unmanned underwater vehicle, the TWSA can estimate 3D trajectories assisted by a trained deep learning model, with a root mean square error of ≈0.02. Thus, the TWSA‐based assisted perception holds immense potential for enhancing unmanned underwater vehicle near‐field perception and navigation capabilities across a wide range of applications.

Optimizing underwater connectivity through multi-attribute decision-making for underwater IoT deployments using remote sensing technologies

The underwater Internet of Things (UIoT) and remote sensing are significant for biodiversity preservation, environmental protection, national security, disaster assistance, and technological innovation. Assigning tasks to autonomous underwater vehicles (AUVs) is a fundamental challenge in underwater technology and exploration. Remote sensing and AUVs are vital for pollution detection, disaster prevention, marine observation, and ocean monitoring. This work presents an optimized network connectivity using a multi-attribute decision-making approach for underwater IoT deployment. A feature engineering approach highlights the significant characteristics of underwater things, incorporating remote sensing data, and a multi-objective optimization method is used to select optimal UIoT for effective task allocation in deep-sea environments. A balance between data transmission, energy economy, and operational performance is necessary for efficient task distribution. Effective communication algorithms and protocols are needed to maintain environmental sustainability, protect marine ecosystems, and improve underwater monitoring enhanced by remote sensing technologies. Multi-criteria decision-making (MCDM) is beneficial for addressing various challenges in underwater technology, considering factors such as mission objectives, energy efficiency, environmental conditions, vehicle performance, safety, and much more. The proposed criteria importance through intercriteria correlation (CRITIC) methodology will assess technical competencies like communication, resilience, navigation, and safety in an underwater environment, leveraging remote sensing and aiding decision-makers in selecting appropriate undersea devices and vehicles for enhancing communication and transportation. This method prioritizes characteristics and aligns them with specific objectives, improving decision-making quality in the marine environment.

In-motion fine alignment algorithm for AUV based on improved extended state observer and Kalman filter

Abstract Achieving fast initial alignment has always been a major challenge for autonomous underwater vehicle navigation, especially in motion. During the fine alignment process, traditional algorithms often face the slow convergence of the azimuth misalignment angle. To solve this problem, this paper proposes a fast alignment algorithm based on the improved extended state observer (ESO) and Kalman filter. Specifically, since the application potential of ESO in fine alignment has not been fully explored by previous studies, this paper introduces the improved ESO through the in-depth research of traditional ESO. The obtained improved ESO is then combined with Kalman filter to constitute the proposed fast alignment algorithm. The algorithm first uses Kalman filter to estimate the two horizontal misalignment angles. Once the horizontal angles estimation of Kalman filter tends to be stable, the improved ESO is used to estimate the azimuth misalignment angle. Simulation and lake test show the proposed algorithm has excellent performance of in-motion alignment. The comparative analysis shows that the algorithm significantly improves the convergence speed of azimuth alignment angle.

Research on Model Reduction of AUV Underwater Support Platform Based on Digital Twin

Digital twin technology, as a data-driven and model-driven innovation means, plays a crucial role in the process of digital transformation and intelligent upgrading of the marine industry, helping the industry to move towards a new stage of more intelligent and efficient development. In order to solve the defects of the Autonomous Underwater Vehicle (AUV) underwater support platform structure deformation field, digital twin technology and model reduction technology are applied to an AUV underwater support platform, and a five-dimensional digital twin model of the AUV underwater support platform is studied, including five dimensions: physical world, digital world, twin data center, service application, and data connection. The digital twin of the subsea support platform is established by using the digital twin modeling technology. The POD method is used to calculate the deformation field matrix of the support structure of the subsea support platform under the 0–5 sea state, and the corresponding eigenvalues and eigenvectors are obtained. By intercepting the eigenvectors corresponding to the eigenvalues of the high energy proportion, the low-order equation is constructed, and the reduced-order model under each sea state can be quickly solved. The experimental results show that the model reduction technology can greatly shorten the model solving time, and the calculated results are highly consistent with the simulation results of the finite element full-order model, which can realize the rapid analysis of the deformation response of the subsea support platform structure, and provide a theoretical basis and technical support for the subsequent simulation, state evaluation, visual monitoring, and predictive maintenance.

Knowledge hierarchy-based dynamic multi-objective optimization method for AUV path planning in cooperative search missions

This study focuses on path planning for Autonomous Underwater Vehicles (AUVs) in underwater cooperative search missions. The complexities of the ocean environment, the uncertainty of target movements, and limited communication capabilities render underwater cooperative search missions dynamic multi-objective optimization challenges. Studies focusing on communication-constrained search strategies still lack reliable metrics for assessing underwater communication quality and do not consider communication dead zones. Addressing these deficiencies, this paper introduces a novel communication optimization strategy utilizing packet error rate as a metric for assessing communication efficacy, complemented by a potential field-based method for navigating out of communication dead zones. To tackle the inherently dynamic nature of cooperative search missions for mobile underwater targets, we propose a dynamic multi-objective optimization algorithm that employs a knowledge hierarchy strategy. This method enhances the NSGA-II algorithm's efficiency by extracting effective gene segments from historical Pareto solution set and generating a new initial population through recombination, hierarchy, and prediction. Distinct from other advanced dynamic multi-objective optimization approaches that are limited to theoretical problems, our approach is directly applicable to practical scenarios. The effectiveness and practicality of the proposed method are validated through a series of simulation experiments considering the impact of underwater acoustic communication. These results demonstrate that this research is not only innovative in theory but also holds significant engineering value and practical prospects in real-world applications.

Kinematic analysis and advanced control of a vectored thruster based on 3RRUR parallel manipulator for micro-size AUVs

Abstract Autonomous underwater vehicles (AUVs) have played a pivotal role in advancing ocean exploration and exploitation. However, traditional AUVs face limitations when executing missions at minimal or near-zero forward velocities due to the ineffectiveness of their control surfaces, considerably constraining their potential applications. To address this challenge, this paper introduces an innovative vectored thruster system based on a 3RRUR parallel manipulator tailored for micro-sized AUVs. The incorporation of a vectored thruster enhances the performance of micro-sized AUVs when operating at minimal and low forward speeds. A comprehensive exploration of the kinematics of the thrust-vectoring mechanism has been undertaken through theoretical analysis and experimental validation. The findings from theoretical analysis and experimental confirmation unequivocally affirm the feasibility of the devised thrust-vectoring mechanism. The precise control of the vector device is studied using Physics-informed Neural Network and Model Predictive Control (PINN-MPC). Through the adoption of this pioneering thrust-vectoring mechanism rooted in the 3RRUR parallel manipulator, AUVs can efficiently and effectively generate the requisite motion for thrust-vectoring propulsion, overcoming the limitations of traditional AUVs and expanding their potential applications across various domains.

Steering characteristics and path following control of a bionic underwater vehicle with multiple locomotion modes

Steering characteristics and path following control of a bionic underwater vehicle with multiple locomotion modes

A dual-stage coverage path planning method for bathymetric survey using an AUV in graph-based SLAM framework considering positioning uncertainty

Autonomous underwater vehicles (AUVs) are becoming increasingly popular for seabed coverage survey tasks but suffer from a serious problem of positioning error accumulation, which leads to low path tracking accuracy and a low coverage rate of the target area. A dual-stage AUV path planning (DSPP) approach was proposed within a simultaneous localization and mapping (SLAM) framework for coverage bathymetric surveys. In the first stage, a global coverage path is designed in a lawnmower style. Crossed paths are added to generate periodic loop closures in SLAM, which can improve the efficiency of online replanning. In the second stage, with a probabilistic neural network (PNN) used for feature-rich bathymetric sub-map identification, an entropy-based online path replanning approach was implemented using 20 templates to adjust the global path to produce accurate loop closures for positioning error correction. Two bathymetric datasets were used to test the proposed DSPP method using a self-developed experimental simulation system. Experiments verified that DSPP can significantly improve the positioning accuracy and coverage rate with acceptably longer paths. In a 6000 m× 5000 marea,the DSPP approach reduced the positioning error by 76% compared with the pure dead reckoning method, and the coverage ratio reached 98.15%.

Autonomous Underwater Vehicle (AUV) Motion Design: Integrated Path Planning and Trajectory Tracking Based on Model Predictive Control (MPC)

This paper attempts to develop a unified model predictive control (MPC) method for integrated path planning and trajectory tracking of autonomous underwater vehicles (AUVs). To deal with the computational burden of online path planning, an event-triggered model predictive control (EMPC) method is introduced by using the environmental change as a triggering mechanism. A collision hazard function utilizing the changing rate of hazard as a triggering threshold is proposed to guarantee safety. We further give an illustration of how to calculate this threshold. Then, a Lyapunov-based model predictive control (LMPC) framework is developed for the AUV to solve the trajectory tracking problem. Leveraging a nonlinear integral sliding mode control strategy, we construct the contraction constraint within the formulated LMPC framework, thereby theoretically ensuring closed-loop stability. We derive the necessary and sufficient conditions for recursive feasibility, which are subsequently used to prove the closed-loop stability of the system. In the simulations, the proposed path planning and tracking control are verified separately and integrated and combined with static and dynamic obstacles.

Integrated analytic hierarchy process and multi-criteria decision-making approach: An Application for Unmanned Underwater Vehicle Control Method Selection

Using Multi-Criteria Decision Analysis (MCDA) techniques, it is possible to thoroughly examine all available options and determine the most desirable ones. However, the prevalence of these methods, which employ various computational changes, can result in disparate outcomes. Therefore, the final rankings should be as dependable and consistent as possible. In this study, the consistency of the findings obtained from the TOPSIS, COPRAS, MAUT, and MOOSRA methodologies was investigated in the real-world problem of Unmanned Underwater Vehicle Control Method Selection by forming decision matrices through simulations.

Self-powered piezoelectric sensing system for rotational speed detection of AUV propellers

To realize high-precision self-supplied speed measurement of Autonomous Underwater Vehicle (AUV) propellers, a self-powered piezoelectric speed sensing and detecting system is developed, which consists of a self-powered piezoelectric detection module and energy management and sensing circuits. The system uses three piezoelectric cantilevers for power and sensing detection. The proposed speed detection system is based on the piezoelectric effect and realizes self-powered sensing by utilizing the kinetic energy generated during the rotation of the propeller into electrical energy. The comprehensive model of the proposed piezoelectric sensing system under periodic toggled excitation is established and analyzed. The optimal output power of the system is investigated through theoretical analysis and simulation. The experimental system of the self-powered speed sensing system is constructed and the effect of speed measurement is tested. Experimental results show that the maximum output power of the piezoelectric self-powered sensing system developed in this work is 218 μW, which is much higher than that of the purely triboelectric sensing system. And the average detection error of the self-powered sensing system is 1.4% and the linearity is 1.09% when the propeller speed ranges from 0 to 300 r/min. Compared with similar self-powered systems, the proposed piezoelectric self-powered system is more advantageous in terms of measurement accuracy and can provide more accurate propeller speed measurement.

On the Development of an Acoustic Image Dataset for Unexploded Ordnance Classification Using Front-Looking Sonar and Transfer Learning Methods.

This research aimed to develop a dataset of acoustic images recorded by a forward-looking sonar mounted on an underwater vehicle, enabling the classification of unexploded ordnances (UXOs) and objects other than unexploded ordnance (nonUXOs). The dataset was obtained using digital twin simulations performed in the Gazebo environment utilizing plugins developed within the DAVE project. It consists of 69,444 sample images of 512 × 399 resolution organized in two classes annotated as UXO and nonUXO. The obtained dataset was then evaluated by state-of-the-art image classification methods using off-the-shelf models and transfer learning techniques. The research included VGG16, ResNet34, ResNet50, ViT, RegNet, and Swin Transformer. Its goal was to define a base rate for the development of other specialized machine learning models. Neural network experiments comprised two stages-pre-training of only the final layers and pre-training of the entire network. The experiments revealed that to obtain high accuracy, it is required to pre-train the entire network, under which condition, all the models achieved comparable performance, reaching 98% balanced accuracy. Surprisingly, the highest accuracy was obtained by the VGG model.

Gaze-Assisted Prescribed Performance Controller for AUV Trajectory Tracking in Time-Varying Currents

Trajectory tracking for underactuated autonomous underwater vehicles (AUVs) is challenging due to coupling dynamics, modeling inaccuracies, and unknown disturbances. To tackle this, we propose a decoupling gaze-assisted prescribed performance controller (GAPPC). We first use an error transformation approach to achieve the prescribed performance, incorporating the line-of-sight (LOS) algorithm and an event-triggering mechanism to handle the kinematic characteristics of underactuated AUVs. Next, we develop a control strategy for the transformed error that does not require knowledge of the model parameters, including fast dynamic compensation to reduce steady-state errors. Finally, we analyze the controller’s stability and present simulation results. Simulations, which account for modeling inaccuracies and unknown ocean currents, show that the GAPPC improves stability errors by 67.3% compared to the adaptive robust controller.

A hybrid RVO-MPPI approach for efficient collision avoidance for multiple autonomous underwater vehicles

Collision-avoidance path planning for Autonomous Underwater Vehicles (AUVs) in complex underwater environments presents a significant challenge, particularly considering the maneuverability limitations of fin-rudder-controlled AUVs. This study proposes a novel navigation approach that combines the strengths of Reciprocal Velocity Obstacle (RVO) principles and Model Predictive Path Integral (MPPI) control to enhance autonomous obstacle avoidance for multiple AUVs. The key innovation lies in leveraging RVO-based obstacle avoidance velocities to enhance both action sampling and cost function optimization in MPPI framework. Specifically, a neural network trained through supervised learning is utilized to convert the RVO-provided avoidance velocities into desired actions for MPPI. Additionally, these velocities are incorporated into the MPPI cost function, thereby enhancing the optimization process and facilitating convergence towards optimal, collision-free navigation solutions. Simulation experiments validated the method's capability to provide effective obstacle avoidance behavior across various complex scenarios. Results indicate that this approach successfully integrates MPPI's predictive capabilities with RVO's velocity avoidance logic, enabling AUVs to navigate efficiently through complex underwater environments while ensuring mission safety and effectiveness. This approach offers a promising new solution for autonomous navigation of AUVs, with potential applications in ocean exploration, underwater search and rescue, and other maritime operations. Future research will explore the performance of this approach in real-world underwater environments and consider the impact of additional dynamic factors.

Maritime ISAR detection and tracking algorithm for multiple maneuvering extended vessels in heavy-tailed clutter using SK-MM-Sub-RMM-MB-TBD filter

Multiple extended objects tracking (EOT) tasks play an important role in computer vision and engineering applications of artificial intelligence. In particular, nonlinear marine object imaging and tracking has received an increasing amount of attention due to its enormous application potential in the field of marine engineering, Autonomous Underwater Vehicles, and Remotely Operated Vehicles. Under high sea state, ship-EOTs perform complex maneuvering movements due to strong disturbances such as sea winds and sea waves. In this paper, we exploit emergent maneuvering EOTs (M-EOTs) methodologies in heavy-tailed clutter. We propose a M-EOT procedure in real-time scenario based on the popular multi-Bernoulli (MB)-TBD filter in maritime inverse synthetic aperture radar (ISAR) systems, and in particular, we describe the extended ship target state through the random matrices model (RMM). In RMM, scatter centers are distributed symmetrically around the M-EOT's centroid. However, in ship M-EOT scenario, the distribution over the whole object is not symmetrical, but distributed and skewed in some portions while a target maneuvers. To solve this problem, a novel robustness observation model is represented by using skewed (SK) non-symmetrically normal distribution and multiple model (MM) MB-TBD with more than one ellipse. Simulation and experimental results illustrate that the proposed SK-MM-Sub-RMM-MB-TBD filter outperforms the existing filters for M-EOTs.

Removing nonrigid refractive distortions for underwater images using an attention-based deep neural network

Refractive distortions in underwater images usually occur when these images are captured through a dynamic refractive water surface, such as unmanned aerial vehicles capturing shallow underwater scenes from the surface of water or autonomous underwater vehicles observing floating platforms in the air. We propose an end-to-end deep neural network for learning to restore real scene images for removing refractive distortions. This network adopts an encoder-decoder architecture with a specially designed attention module. The use of the attention image and the distortion field generated by the proposed deep neural network can restore the exact distorted areas in more detail. Qualitative and quantitative experimental results show that the proposed framework effectively eliminates refractive distortions and refines image details. We also test the proposed framework in practical applications by embedding it into an NVIDIA JETSON TX2 platform, and the results demonstrate the practical value of the proposed framework.

Sim-to-real transfer of adaptive control parameters for AUV stabilisation under current disturbance

Learning-based adaptive control methods hold the potential to empower autonomous agents in mitigating the impact of process variations with minimal human intervention. However, their application to autonomous underwater vehicles (AUVs) has been constrained by two main challenges: (1) the presence of unknown dynamics in the form of sea current disturbances, which cannot be modelled or measured due to limited sensor capability, particularly on smaller low-cost AUVs, and (2) the nonlinearity of AUV tasks, where the controller response at certain operating points must be excessively conservative to meet specifications at other points. Deep Reinforcement Learning (DRL) offers a solution to these challenges by training versatile neural network policies. Nevertheless, the application of DRL algorithms to AUVs has been predominantly limited to simulated environments due to their inherent high sample complexity and the distribution shift problem. This paper introduces a novel approach by combining the Maximum Entropy Deep Reinforcement Learning framework with a classic model-based control architecture to formulate an adaptive controller. In this framework, we propose a Sim-to-Real transfer strategy, incorporating a bio-inspired experience replay mechanism, an enhanced domain randomisation technique, and an evaluation protocol executed on a physical platform. Our experimental assessments demonstrate the effectiveness of this method in learning proficient policies from suboptimal simulated models of the AUV. When transferred to a real-world vehicle, the approach exhibits a control performance three times higher compared to its model-based nonadaptive but optimal counterpart.

Few-shot fault diagnosis of underwater thrusters based on semi-supervised prototypical network with SimAM attention and auxiliary classifier

Autonomous Underwater Vehicles (AUVs) are essential equipment for ocean development, and the propeller is the power component of an AUV. Fault diagnosis of the propeller is crucial to ensure the normal operation of AUVs. However, in practical engineering, the lack of sufficient labeled data poses a challenge for applying deep learning in propeller fault diagnosis. This paper proposes a semi-supervised prototypical network with SimAM attention and auxiliary classifier (SSPN-SimAM-AC) for few-shot fault diagnosis of underwater thrusters. Specifically, we design a rule for label assignment to screen for out-of-distribution samples. The SimAM attention mechanism is added to the feature extractor to enhance the capability of extracting sample features. Using a semi-supervised prototypical network, the prototypes are updated with unlabeled data, and based on this, an auxiliary classifier is designed to achieve better fault identification. The effectiveness of this method is validated using two sets of propeller fault data, and the experimental results demonstrate that the method can diagnose propeller faults of underwater vehicles under small sample conditions.

Experimental and Simulation Study on Flow-Induced Vibration of Underwater Vehicle

At high speeds, flow-induced vibration noise is the main component of underwater vehicle noise. The turbulent fluctuating pressure is the main excitation source of this noise. It can cause vibration of the underwater vehicle’s shell and eventually radiate noise outward. Therefore, by reducing the turbulent pressure fluctuation or controlling the vibration of the underwater vehicle’s shell, the radiation noise of the underwater vehicle can be effectively reduced. This study designs a cone–column–sphere composite structure. Firstly, the effect of fluid–structure coupling on pulsating pressure is studied. Next, a machine learning method is used to predict the turbulent pressure fluctuations and the fluid-induced vibration response of the structure at different speeds. The results were compared with experimental and numerical simulation results. The results show that the deformation of the structure will affect the flow field distribution and pulsating pressure of the cylindrical section. The machine learning method based on the BP (back propagation) neural network model can quickly predict the pulsating pressure and vibration response of the cone–cylinder–sphere composite structure under different Reynolds numbers. Compared with the experimental results, the error of the machine learning prediction results is less than 7%. The research method proposed in this paper provides a new solution for the rapid prediction and control of hydrodynamic vibration noise of underwater vehicles.

Design a Computer Controlled Unmanned Underwater Vehicle Thruster Using Smart Closed Loop Controller

The first step of designing an Unmanned Underwater Vehicle (UUV) depends mainly on the selection of the thrusters as UUVs require many thrusters to be installed onboard for smooth operation. The widely used thrusters in UUVs are has relatively large dimensions and they just spin in the water without reporting any kind of important information such as the status of the thruster and the RPM of the motor. This project aims to design and implement a new computer-controlled thrusting system from end-to-end that fits various UUVs with more advantages including smaller size thruster, cheaper in price, and with more additional smart capabilities that reports thruster’s instantaneous RPM, power consumption, proper operation, malfunctions, damages, and water stream blockage.