首页> 外文会议>International Conference on OCEANS MTS/IEEE Washington >ICT application of real-time monitoring and estimation system for set-net fishery
【24h】

ICT application of real-time monitoring and estimation system for set-net fishery

机译:ICT应用实时监测和估算系统对套餐渔业

获取原文

摘要

In this paper, we propose an ICT application of real-time monitoring system to estimate the catch amount within set-net. Generally, set-net fishermen do not know the catch amount in advance before arriving to set-net area, in which they tend to unable to perform the fishing operation effectively. It means that the fishermen might suffer a loss of time, gasoline and labor cost, along with extra effort. In order to support the fishermen to avoid such condition, in our previous study, we have presented the real-time monitoring system based on ICT aided echo sounder for the set-net fishing. Besides, the algorithm based on multiple regression analysis (MRA) for estimation of catch amount was proposed so that fishermen could predict the catch amount before they leave the port for harvesting. However, the catch estimation was not sufficiently accurate in practical application. The reason of this is because the daily catch amount varies from some few tons to 250 tons, whereas statistics of ping intensity do not vary so much, thus, it is not easy to build a model of MRA adequately. This study aims to develop a more accurate catch estimation algorithm based on our previous study. The experiment of this study is conducted in Hokkaido and Toyama Prefecture. To transform the catch record appropriately for MRA, Box-Cox transformation (BCT) as pre-processing transformation is applied in this paper. The catch estimation has been done based on fishermen's empirical estimation through a pre-processing transformation and statistical analysis of reflection data in each layer depth. The study showed that the present system could improve the estimation accuracy. In order to measure the differences between catch estimation by the developed algorithm and the real catch record, Relative Absolute Error (RAE) is used in this study. The RAE for Hokkaido experimental site was 1% and in Toyama is 22%, respectively. The results show that the estimation accuracy of the present method is sufficie- tly adequate for practical application.
机译:在本文中,我们提出了实时监测系统的ICT应用来估计设定网络内的捕获量。一般来说,在到达设定网面积之前,设定净渔民在抵达之前,不知道捕获量,其中他们倾向于有效地无法执行钓鱼操作。这意味着渔民可能会损失时间,汽油和劳动力成本以及额外的努力。为了支持渔民避免这种情况,在我们以前的研究中,我们介绍了基于ICT辅助回声发声器的实时监控系统,用于设定净捕鱼。此外,提出了基于多元回归分析(MRA)的算法,以估计捕获量,使渔民可以预测捕获量,然后才能离开端口进行收获。然而,在实际应用中,捕获估计并不充分准确。原因是因为每日捕捞量因几吨而异到250吨,而Ping强度的统计数据不会变化,因此,充分构建MRA模型并不容易。本研究旨在基于我们以前的研究开发更准确的捕获估计算法。本研究的实验是在北海道和富山县进行的。为了改变MRA的捕获记录,在本文中应用了作为预处理变换的Box-Cox转换(BCT)。通过预处理变换和每层深度中的反射数据的统计分析,基于渔民的实证分析来完成捕获估计。该研究表明,本系统可以提高估计精度。为了测量发达的算法捕获估计之间的差异和实际捕获记录,本研究使用相对绝对误差(RAE)。北海道实验部位的RAE分别为1%,富山分别为22%。结果表明,目前方法的估计精度适当适用于实际应用。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号