To solve the problem of active collision avoidance for unmanned surface vehicles (USVs) in a complex maritime environment, method deep reinforcement learning (DRL) based on proximal policy optimization (PPO) is proposed. master by self-learning, mathematical model USV, dynamic obstacle generation model, and reward mechanism were established. Using relative position between USV information closest point approach (CPA), high-dimensional state features including track layer threat constructed....

Abstract In view of the deficiencies in traditional visual water surface object detection, such as existence non-detection zones, failure to acquire global information, and a single-shot multibox detector (SSD) detection algorithm remote low precision small objects, this study proposes from panoramic vision based on an improved SSD. We reconstruct backbone network for SSD algorithm, replace VVG16 with ResNet-50 network, add five layers feature extraction. More abundant semantic information...

This paper takes the development of commercial paragliding clubs in Hubei Province as research object, and studies some common problems operation management clubs, which mainly include: unclear overall positioning business objectives; weak consumption ability club members; talent shortage; lack in-depth understanding characteristics service industry club. On basis interviews questionnaires, this uses mathematical statistics to analyze results survey, puts forward suggestions on future...

It has gained much attention since the birth of multi-source information fusion technology for its excellent overall system performance. This paper introduces principle and extends model structure target fusion, methods practice problems such as detection position identification STA sensor control are summarized. Finally according to complex environment, topology, network bandwidth other practical requirements in USV colony, it provides support The relevant research directions given,...

For the problem of local equilibrium point in path planning, a smooth escape algorithm is proposed. In this method, confirmation rule for obstructing obstacles established to confirm routes an unmanned surface vehicle and causing from multiple detecting targets. A model tracking constructed guide bypass safely while points. switching ensure change heading planned by switching. The idea ‘wall-following’ curve smoothing are integrated then planning method escaping given. Simulation results...

To find the law of ship seismic wave field in shallow sea, three-dimensional staggered grid method is adopted. Firstly, this algorithm used to compare fields YOZ and XOZ planes, which shows that t characteristics them are basically same. Secondly, it change source depth, closer near seabed, stronger surface field; energy ratio focus horizontal direction higher, while vertical higher. The lower centre frequency is, signal. Last, far-field observed, P S merge into one group, velocity somewhere...

To find the law of ship seismic wave field in shallow sea, three-dimensional staggered grid method is adopted. Firstly, this algorithm used to compare fields YOZ and XOZ planes, which shows that t characteristics them are basically same. Secondly, it change source depth, closer near seabed, stronger surface field; energy ratio focus horizontal direction higher, while vertical higher. The lower centre frequency is, signal. Last, far-field observed, P S merge into one group, velocity somewhere...

There are some problems in the training of paragliding talents currently China: direction is too single, resulting lack comprehensive talents; structure coaches and management system not perfect enough to meet needs international competitive field; guarantee high-level athletes enough. Facing on these problems, this paper adopts methods literature review, investigation in-depth interviews analyze four modes national team training, college club “individual-family” mode. The puts forward...