圆形重力式深水网箱受灾破坏机理分析

我国沿海地区台风等海洋灾害性气候高发,重力式深水网箱易受到台风等恶劣海洋环境的严重威胁,严重阻碍了我国网箱养殖产业发展。重力式深水网箱受灾破坏形式主要包括浮架受损、网衣严重变形和锚碇脱锚。以温州南麂岛海域的网箱设施情况和水文资料为样本,选取了周长分别为40,60 m及80 m的圆形重力式深水网箱,在水深20,22 m及23 m,波高4~10 m,周期5.4~12.5 s时,通过计算网箱在波浪作用下的水动力响应,利用BP神经网络构建受灾破坏因子与结构受损的定量关系,并利用灰色关联度识别受灾破坏主控因子。分析结果表明,BP神经网络方法可以很好地拟合自变量(受灾破坏因子)与应变量(结构受损)的定量关系。在锚绳破坏中,波高的关联系数为0.59,波高为主控因子;在网衣变形中,周长的关联数为0.39,周长为主控因子;在浮架受损中,周长的关联数为0.58,周长为主控因子。

A study on failure mechanism of circular gravity net cage

The East China Sea suffers from high frequent typhoon, storm surge and other marine disastrous climates, which seriously hinder the healthy development of the net cage aquaculture industry in China. The failures of the offshore deep gravity net cage mainly include the damage of the floating collar, deformation of the net and separation of the anchor. Based on the survey and the hydrological data of the Nanji Island in Wenzhou, high density polyethylene (HDPE) gravity cage with the circumference of 40, 60 and 80 m are chosen as the typical facilities in this study. In terms of water depth at 20, 22 and 23 m, wave height ranging 4~10 m and wave duration ranging 5.4~12.5 s, the dynamic response simulation of the cage are conducted. Based on the simulated results, the quantitative relationship between hazard degree and damaging factors of the cage is established by BP (Back Propagation) neural network and the controlling hazard factor is identified by grey correlation analytical method. The results show that BP neural network method can fit well the quantitative relationship between independent variables (hazard degree) and dependent variables (damaging factors of net cage); for the failure of the anchor rope, the correlation coefficient of wave height is 0.59 and wave height is a dominant factor; for the deformation of the net and the damage of floating collar, the correlation coefficients of circumference are 0.39 and 0.58 respectively and circumference is a dominant factor.