This paper proposes an improved TD3 (Twin Delayed Deep Deterministic Policy Gradient) algorithm to address the flaws of low success rate and slow training speed, when using the original TD3 algorithm in mobile robot path planning in dynamic environment. Firstly, prioritized experience replay and transfer learning are introduced to enhance the learning efficiency, where the probability of beneficial experiences being sampled in the experience pool is increased, and the pre-trained model is applied in an obstacle-free environment as the initial model for training in a dynamic environment. Secondly, dynamic delay update strategy is devised and OU noise is added to improve the success rate of path planning, where the probability of missing high-quality value estimate is reduced through changing the delay update interval dynamically, and the correlated exploration of the mobile robot inertial navigation system in the dynamic environment is temporally improved. The algorithm is tested by simulation where the Turtlebot3 robot model as a training object, the ROS melodic operating system and Gazebo simulation software as an experimental environment. Meanwhile, the result shows that the improved TD3 algorithm has a 16.6 % increase in success rate and a 23.5 % reduction in algorithm training time. A generalization experiment was designed finally, and it indicates that superior generation performance has been acquired in mobile robot path planning with continuous action spaces through the improved TD3 algorithm.

The visual signal object detection technology of deep learning, as a high-precision perception technology, can be adopted in various image analysis applications, and it has important application prospects in the utilization and protection of marine biological resources. While the marine environment is generally far from cities where the rich computing power in cities cannot be utilized, deploying models on mobile edge devices is an efficient solution. However, because of computing resource limitations on edge devices, the workload of performing deep learning-based computationally intensive object detection on mobile edge devices is often insufficient in meeting high-precision and low-latency requirements. To address the problem of insufficient computing resources, this paper proposes a lightweight process based on a neural structure search and knowledge distillation using deep learning YOLOv8 as the baseline model. Firstly, the neural structure search algorithm was used to compress the YOLOv8 model and reduce its computational complexity. Secondly, a new knowledge distillation architecture was designed, which distills the detection head output layer and NECK feature layer to compensate for the accuracy loss caused by model reduction. When compared to YOLOv8n, the computational complexity of the lightweight model optimized in this study (in terms of floating point operations (FLOPs)) was 7.4 Gflops, which indicated a reduction of 1.3 Gflops. The multiply–accumulate operations (MACs) stood at 2.72 G, thereby illustrating a decrease of 32%; this saw an increase in the AP50, AP75, and mAP by 2.0%, 3.0%, and 1.9%, respectively. Finally, this paper designed an edge computing service architecture, and it deployed the model on the Jetson Xavier NX platform through TensorRT.

The formation operation of autonomous underwater vehicles (AUVs) plays a significant role in advancing the mission success of Internet of Underwater Things (IoUT). Nevertheless, potential noncooperative goals and scarce underwater acoustic communication resources pose challenges in the formation deployment of AUVs. Regarding this, this article develops a noncooperative game-based dynamic formation strategy for multiple AUVs with event-driven communications. Initially, the cost evaluation framework is established to formulate the noncooperative dynamic formation game of multiple AUVs. Then, to solve this game issue, a gradient-based distributed control algorithm is developed to steer AUVs to achieve the nash equilibrium (NE). Meanwhile, to optimize communication resource consumption and avoid impractical communication intervals, a hybrid dynamic event-driven communication mechanism (HDECM) is constructed to manage interactions among vehicles. The developed HDECM possesses two distinctive features compared to conventional approaches: 1) preformulating system-stability-ensured silent intervals; and 2) reducing redundant communications. Finally, the stability analysis and semi-physical simulations are conducted to illustrate the effectiveness and superiority of the proposed scheme.

Marine biological object detection is of great significance for the exploration and protection of underwater resources. There have been some achievements in visual inspection for specific objects based on machine learning. However, owing to the complex imaging environment, some problems, such as low accuracy and poor real-time performance, have appeared in these object detection methods. To solve these problems, this paper proposes a detection method of marine biological objects based on image enhancement and YOLOv5S. Contrast-limited adaptive histogram equalization is taken to solve the problems of underwater image distortion and blur, and we put forward an improved YOLOv5S to improve accuracy and real-time performance of object detection. Compared with YOLOv5S, coordinate attention and adaptive spatial feature fusion are added in the improved YOLOv5S, which can accurately locate the target of interest and fully fuse the features of different scales. In addition, soft non-maximum suppression is adopted to replace non-maximum suppression for the improvement of the detection ability for overlapping objects. The experimental results show that the contrast-limited adaptive histogram equalization algorithm can effectively improve the underwater image quality and the detection accuracy. Compared with the original model (YOLOv5S), the proposed algorithm has a higher detection accuracy. The detection accuracy AP50 reaches 94.9% and the detection speed is 82 frames per second; therefore, the real-time performance can be said to reach a high level.

The fault detection and isolation (FDI) accuracy directly affects the redundant inertial measurement unit (RIMU) working state. The previous methods are difficult to apply to the four-axis system because they do not have enough information to allow fault isolation. Therefore, in this paper, an extended RIMU system is proposed to implement fault location in a four-axis system. The extended system consists of three additional accelerometers and a four-axis RIMU. The specific force information in the gyro-free strapdown inertial navigation system (GFSINS) is used to obtain reference angular velocity information. Then, based on the extended system, a fault detection procedure is designed to achieve faulty gyro isolation and system reconfiguration. Simulation results show that the extended system can effectively isolate gyroscope failures in a four-axis RIMU system. Compared to the non-redundant three-axis system, the extended system is 1.75 times more reliable and has 37% better average navigation accuracy.

To address the challenge of suboptimal object detection outcomes stemming from the deformability of marine flexible biological entities, this study introduces an algorithm tailored for detecting marine flexible biological targets. Initially, we compiled a dataset comprising marine flexible biological subjects and developed a Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm, supplemented with a boundary detection enhancement module, to refine underwater image quality and accentuate the distinction between the images’ foregrounds and backgrounds. This enhancement mitigates the issue of foreground-background similarity encountered in detecting marine flexible biological entities. Moreover, the proposed adaptation incorporates a Deformable Convolutional Network (DCN) network module in lieu of the C2f module within the YOLOv8n algorithm framework, thereby augmenting the model’s proficiency in capturing geometric transformations and concentrating on pivotal areas. The Neck network module is enhanced with the RepBi-PAN architecture, bolstering its capability to amalgamate and emphasize essential characteristics of flexible biological targets. To advance the model’s feature information processing efficiency, we integrated the SimAM attention mechanism. Finally, to diminish the adverse effects of inferior-quality labels within the dataset, we advocate the use of WIoU (Wise-IoU) as a bounding box loss function, which serves to refine the anchor boxes’ quality assessment. Simulation experiments show that, in comparison to the conventional YOLOv8n algorithm, our method markedly elevates the precision of marine flexible biological target detection.