基于权重分析和GAM模型的秋刀鱼舷提网作业性能影响因素

Factors influencing the stick-held net status of Pacific saury (Cololabis saira) fishery based on weighted analysis and GAM

摘要:

根据2016年7—10月和2017年6—10月蓬莱京鲁渔业有限公司“鲁蓬远渔019”在西北太平洋进行的海上秋刀鱼舷提网网具性能测试试验中收集的作业参数、网具深度以及不同水层水流速度等信息,结合提升回归树模型(boosting regression tree, BRT)权重分析结果,基于广义加性模型(generalized additive model, GAM)探讨各因素对舷提网网具作业性能的影响,分析影响因子与舷提网主要性能参数(最大沉降深度和提升速度)的关系。结果表明:影响网具最大沉降深度的因素中,权重在前4位的为30 m水层流速(20.15%)、60 m水层流速(18.92%)、下纲松放长度(16.85%)和10 m水层流速(15.52%);影响提升速度的前4位影响因子为绞网速度(23.17%)、30 m水层流速(20.05%)、10 m水层流速(18.27%)以及60 m水层流速(16.26%)。30 m水层流速、60 m水层流速以及下纲松放长度对网衣最大沉降深度的影响显著,网衣最大沉降深度与各水层流速呈负相关关系,与下纲松放长度呈正相关关系;绞网速度和水层流速(10、30和60 m)均显著影响提升速度,提升速度与绞网速度呈正相关关系,与水层流速呈负相关关系,绞网速度是影响网具提升速度最重要的因素,其次是30 m水层流速、10 m水层流速和60 m水层流速。

Abstract:

Based on the data collected by saury stick-held operation in the northwest Pacific from July to October in 2016 and from June to October in 2017, we analyzed the factors(fishing operations and marine environment) influencing the stick-held net status of Pacific saury(Cololabis saira) fishery based on boosting regression tree weighted analysis and generalized additive model. The results showed that: Among the factors influencing the maximum net sinking depth, the top four factors with weight are current speed of 30 m layer(20.15%), current speed of 60 m layer(18.92%), length of lead line loose(16.85%) and current speed of 10 m layer(15.52%); The top four influencing factors of lifting speed were hauling speed(23.17%), current speed of 30 m layer(20.05%), current speed of 10 m layer(18.27%) and current speed of 60 m layer(16.26%). Current speed of 30 m and 60 m layers and length of lead line loose had significant effects on the maximum net sinking depth;The maximum net sinking depth was positively correlated with the length of lead line loose, there was a negative correlation between the maximum net sinking depth and current speed of 10 m, 30 m and 60 m layers; hauling speed and current speed of 10 m, 30 m and 60 m layers had significant effects on lifting speed; the lifting speed was positively correlated with the hauling speed; there was a negative correlation between the lifting speed and current water speed. Hauling speed was the most important factor affecting the lifting speed, followed by current speed of 30 m layer, current speed of 10 m layer and current speed of 60 m layer.