Autonomous Underwater Vehicle Propulsion Research Articles

Squeeze‐and‐excitation attention residual learning of propulsion fault features for diagnosing autonomous underwater vehicles

AbstractGiven the demanding and unpredictable operational conditions, autonomous underwater vehicles (AUVs) often encounter different propulsion faults, leading to significant economic losses and mission impairments. To address this challenge, vibratory time‐series features can be extracted for the precise propulsion fault diagnosis of AUVs. A squeeze‐and‐excitation (SE) attention residual network (SEResNet) is therefore put forward to enhance the feature extraction for AUV propulsion fault diagnosis. By leveraging the vibratory time‐series data obtained from the AUV, an SE attention mechanism is embedded into a residual network. This integration facilitates the extraction of pertinent vibratory fault features, subsequently utilized for accurate diagnosis of any propulsion faults. The effectiveness of the proposed SEResNet was validated through its application to an actual experimental AUV, with comparison against the state‐of‐the‐arts. The results reveal that the present SEResNet outperforms all other comparison methods in terms of diagnosis performance for AUV propulsion faults.

A GPU-accelerated 3dBEM supporting the optimization of actively morphing flapping-foil thrusters with application to AUV propulsion

In recent years, flapping thrusters have been getting attention as an eco-friendly alternative to conventional propellers for autonomous underwater vehicles (AUVs) mainly due to their low-frequency operation, advanced maneuverability, and high efficiency. Their operating principle mimics the thrust-producing kinematics found among efficient swimmers in the aquatic environment. The present work focuses on the analysis and optimization of flapping systems with active morphing features for increased performance. The two optimization studies refer to a realistic AUV propulsion scenario for a wing thruster with active hydrofoil-section adjustment, spanwise bend and twist morphing capabilities. A gradient descent approach based on sequential quadratic programming (SQP) is used for optimal tuning of design variables (planform shape, kinematics) targeting efficiency maximization under thrust and effective angle of attack constraints. For the evaluation of each candidate thruster the unsteady boundary element solver 3dBEM is developed and used for wing performance prediction. The solver exploits GPU parallel computation techniques to reduce computational time and enable fast optimization. Results indicate that the optimal thruster with chordline/spanwise bending is 25% more efficient, whereas the same result for the wing with spanwise bending and twist profiles is up to 8.8%. The 3dBEM solver can support the preliminary design process of morphing flapping-foil thrusters.

Investigation of self-adjusting cooling system for the autonomous underwater vehicle propulsion motor

Permanent magnet (PM) motors are widely employed in the autonomous underwater vehicle (AUV) propulsion because of their high efficiency and torque density. Since the AUV housings are sealed and narrow, motors are vulnerable to the risk of overheating, leading to dissolution of the winding insulation and demagnetization of the permanent magnets. In this study, a self-adjusting cooling scheme was proposed, which uses the hydrodynamic pressure to pump seawater into the cooling channel automatically. Firstly, the mathematical model of the flow field and heat transfer was established using equation derivations. Subsequently, to clarify the characteristics of the self-adjusting cooling system, the effect of the AUV speed on the thermal and fluid characteristics was investigated using computational fluid dynamics. Results show that the flow rate in the channel increases from 2 to 8.1 L/min as the AUV speed increases from 15 to 50 kn, resulting in the improvement of the heat transfer coefficient from 41 to 764 W/(m2∙K). Subsequently, the influences of inlet and outlet on the thermal characteristics were studied and found to be beneficial for improving the cooling performance. To investigate the spiral flow channels, the distribution coefficient kC and shape factor Rf were defined. Compared to the closed-cycle cooling system for the AUV propulsion motor, the proposed cooling system eliminated auxiliary devices such as pumps and heat exchangers, which makes it more compact and efficient for AUVs. However, introducing seawater into the AUV housing poses a risk of liquid leakage.

Mathematical modeling of autonomous underwater vehicle propulsion and steering complex operation in oblique (beveled) water flow

The research of nonlinear hydrodynamic characteristics of the propulsion and steering complex (PSC), which influence the accuracy of the plane trajectory motion of an autonomous underwater vehicle (AUV), is carried out. During the underwater vehicle curvilinear motion, its PSC operates in an oblique incident water flow. This leads to a decrease in the PSC thrust force and negatively affects the controlled trajectory motion of the underwater vehicle. The research was conducted for a specific type of AUV for the plane curvilinear motion mode. The mathematical modeling method was chosen as the research method. To this end, the well-known AUV motion mathematical model is supplemented by the control system that simulates (mimics) its trajectory motion. The developed model consists of four main units: an AUV improved model; the vehicle speed setting unit; the nozzle rotation angle control unit; the unit containing the AUV pre-prepared motion trajectories. The research results of the AUV hydrodynamic parameters for several typical trajectories of its motion are presented. The investigated parameters include the following: the required nozzle rotation angle; the vehicle actual motion trajectory; the vehicle velocity; the propeller shaft moment; the propeller thrust force. As a result of the conducted researches, the dependence diagram of the propeller thrust force on the AUV nozzle rotation angle in the speed range from 0.2 m/s to 1 m/s and during the nozzle rotation in the range of up to 35° was constructed. A three-dimensional matrix, which describes the dependence of the propeller thrust force on the incident water flow angle and velocity of the vehicle, was created. The obtained dependence can be used in the synthesis of automatic control systems regulators of AUV plane manoeuvering (shunting) motion of increased accuracy.

Noise cancellation for an autonomous underwater vehicle-towed thin line array through recursive adaptive filtering

The digital thin line array (DTLA) developed at ARL, National University of Singapore have been integrated and tested with different autonomous underwater vehicles (AUV) and tested in the field to detect, localize and track underwater targets. For tow speeds not exceeding 4knots, it has been found that the flow noise is not a major contributor to performance reduction. However, noise generated by the propulsion system of some of the AUVs presents a major interferer to the detection capabilities of DTLA, especially, at low ambient noise conditions. The method proposed in the literature for platform noise cancellation employs a single adaptive filter. The number of the filer taps required in this case would be more than thousands due to the multi-path nature of the AUV noise in the environment of interest and the solution is difficult to converge. To mitigate the above problem, we propose a recursively adaptive strategy, which employs the least mean-square (LMS) algorithm. Our strategy is based on the fact that the underwater channel is sparse for the DTLA and using a reference sensor close to the AUV propulsion system. The noise peaks are eliminated through recursive adaptive filter operations after cross correlating the noise signal at each sensor with the reference. The proposed method has been verified through simulations and we propose to apply this technique in a practical application.

Biomechanical model of batoid (skates and rays) pectoral fins predicts the influence of skeletal structure on fin kinematics: implications for bio-inspired design

Growing interest in the development of bio-inspired autonomous underwater vehicles (AUVs) has motivated research in understanding the mechanisms behind the propulsion systems of marine animals. For example, the locomotive behavior of rays (Batoidea) by movement of the pectoral fins is of particular interest due to their superior performance characteristics over contemporary AUV propulsion systems. To better understand the mechanics of pectoral fin propulsion, this paper introduces a biomechanical model that simulates how batoid skeletal structures function to achieve the swimming locomotion observed in nature. Two rays were studied, Dasyatis sabina (Atlantic ray), and Rhinoptera bonasus (cownose ray). These species were selected because they exhibit very different swimming styles (undulation versus oscillation), but all use primarily their pectoral fins for propulsion (unlike electric rays or guitarfishes). Computerized tomography scans of each species were taken to image the underlying structure, which reveal a complex system of cartilaginous joints and linkages. Data collected from these images were used to quantify the complete skeletal morphometry of each batoid fin. Morphological differences were identified in the internal cartilage arrangement between each species including variations in the orientation of the skeletal elements, or radials, and the joint patterns between them, called the inter-radial joint pattern. These data were used as the primary input into the biomechanical model to couple a given ray skeletal structure with various swimming motions. A key output of the model is an estimation of the uniaxial strain that develops in the skeletal connective tissue in order for the structure to achieve motions observed during swimming. Tensile load tests of this connective tissue were conducted to further investigate the implications of the material strain predictions. The model also demonstrates that changes in the skeletal architecture (e.g., joint positioning) will effect fin deformation characteristics. Ultimately, the results of this study can be used to guide the design of optimally performing bio-inspired AUVs.

Applying the self-tuning fuzzy control with the image detection technique on the obstacle-avoidance for autonomous underwater vehicles

This study suggests a simple searching technique for obstacle-avoidance of autonomous underwater vehicles (AUVs) in varying ocean environments by using the self-tuning fuzzy controller. The corresponding hydrodynamic coefficients for the AUV are obtained by the test of Planar Motion Mechanism (PMM), which serves as the important data inputs for the control system. Subsequently, the self-tuning fuzzy controller would be adopted to command the propulsion of AUVs. The function of obstacle-avoidance is based on the underwater image detection method with the BK triangle sub-product of fuzzy relations which can evaluate the safety and remoteness of the candidate routes and the successive optimal heading of strategic routing can then be selected. In the present simulations, four types of motion control factors are selected as the platform to investigate the maneuvering performance of obstacle-avoidance, i.e. self-tuning control, visibility, safety and current effect. Eventually, the present study indicates that the self-tuning fuzzy controller, combined with the image detection technique based on BK triangle sub-product of fuzzy relations, is verified to be a useful searching technique for obstacle-avoidance of AUVs.

A review of developments towards biologically inspired propulsion systems for autonomous underwater vehicles

Autonomous underwater vehicles (AUVs) are playing an ever-growing role in modern subocean operations, generating a demand for faster, more manoeuvrable designs capable of deployments of increasingly longer durations. In order to meet these demands, vehicle developers have been looking to biological aquatic animals for inspiration. After evolving for millions of years, fish and cetaceans have developed fast efficient locomotion techniques, with levels of manoeuvrability that far outperform conventional engineered marine locomotion systems. This paper aims to give a brief introduction into some of the biologically inspired propulsion mechanisms that have been developed, to explain their strengths, their weaknesses, and the motivation behind them, and then finally to predict future trends in biomimetic AUV propulsion design.

SUKF-based Modeling for General Behavior of the Environment and Its Application in Long-range AUV Propulsion System

Due to the poor adaptive capacity of long-range autonomous underwater vehicles(AUV) for general behavior of the environment,a square-root unscented kalman filter(SUKF) based modeling is proposed.In the offline model of general behavior of the environment,some parameters are time-varying,or not possible to be predicted,these parameters must be estimated and predicted through the information detected by the sensors.According to the sensor's information online and model offline,SUKF is used to estimate the state and parameter about the general behavior of the environment in real time.Modeling for behavior environment of AUV is the main work,taking propulsion system as an example,a model structure of fault used actuator effectiveness factors(AEFs) is constructed,then the SUKF is used for on-line joint estimation of both the states and AEFs parameters of AUV.Simulators are conducted by using the model of long-range AUV,and the results show the validity of the algorithm.Finally,the factors that affect the performance for estimation of SUKF are analyzed.

Decreasing the Radiated Acoustic and Vibration Noise of a Mid-Size AUV

An Odyssey IIb autonomous underwater vehicle (AUV) made by Bluefin Robotics, Inc., was acquired by the Marine Physical Laboratory, Scripps Institution of Oceanography, to conduct research in underwater acoustics as well as provide a platform for other scientific studies. The original Odyssey IIb tail cone was replaced with a ducted fan, vectored thrust system installed on vehicles currently sold by Bluefin. In initial sea tests with the new thrust system, the acoustic self noise levels of the vehicle while underway were 20 to 50 dB higher than typical ocean background noise levels, preventing the vehicle's use as a receiver of low level sounds. Controlled tests were performed to characterize the radiated and vibration noise of the AUV propulsion and actuators. Once this baseline was established, changes were made, mostly to the tail cone propulsion, to decrease the vehicle's self noise. The resulting self noise levels of the AUV from 10 Hz up to 10 kHz measured while underway by a hydrophone mounted on the AUV's inner shroud now are at or below typical shallow water background noise levels except in three bands; below 250 Hz, around 500 Hz, and from 0.9 to 2.0 kHz. The goal of this paper is to describe these changes and their effects in lowering vehicle noise levels.

Guidance and control of a biomimetic autonomous underwater vehicle using body-fin propulsion

Autonomous underwater vehicles (AUVs) with rigid hulls and powered by rotary propellers have problems such as inefficient propulsion, difficulties associated with positioning, limited turning agility and imprecise hovering. Recently, research and development into biomimetic AUVs (BAUVs) have accelerated in order to solve the problems of the propulsion of AUVs by propellers. A testbed BAUV is being developed. A local control law that coordinates body angles by periodically alternating the position of the centre-of-mass in local coordinates is developed. Three parameters are found to suffice to coordinate joint angles in carangiform swimming. A global control law is then developed for solving the waypoint tracking problem. The effectiveness of this method is evaluated using computer simulation. The control performance of the BAUV system with model uncertainties for waypoint tracking under the influence of disturbances is also discussed.