全隐式算法和虚拟神经晶格在延绳钓建模中的应用.docVIP

  Application of whole-implicit algorithm and virtual neural lattice in pelagic longline modeling# ZHANG Xinfeng, CAO Daomei, SONG Liming** 5 10 15 20 25 30 35 (College of Marine Sciences, Shanghai Ocean University, ShangHai 201306) Abstract: This study developed a whole-implicit algorithm and virtual neural lattice to model and simulate the dynamics of pelagic longline gear. The programming, model calculation and behavior simulation of longline dynamics were implemented in R language. Equilibrium state and stable solution of the models were reached. The hooks were more stable in the deepest water layer than that in the midwater layer. It indicated that the whole-implicit method and virtual neural lattice were high efficient and accurate in solving the complicated engineering equations of longline gear, also in the presentation of the motion of longline in sea water. Keywords: whole-implicit algorithm; lattice structure; longline; simulation; modeling; neural network 0 Introduction Longline, a kind of pelagic fishing gear used in the deep ocean, is often used to catch tuna-like fishes such as bigeye tuna, yellowfin tuna and albacore tuna. Mizuno et al. estimated three-dimension shapes of the mainline of longline gear using observed data monitored by an Acoustic Doppler Current Profiler [1]. Miyamoto et al. obtained three-dimension real time measurements of the depth and position of longline using an ultrasonic positioning system [2]. For the difficulties and complex of algorithm and computation in solving engineering system equations, studies using mechanical models to simulate the dynamical process of longline gear were rare to see in recent years. Similar studies on other fishing gears, such as modeling the purse seine and trawl, have appeared recently. Lee et al. modeled and simulated the trawl using physical models [3, 4]. Kim et al. developed a simulation tool using C language to model the purse seine gear [5]. Zhang et al. visualized the part behavior of