上海海洋大学学报2024，Vol.33Issue(6)：1357-1368,12.DOI:10.12024/jsou.20231004323
基于机器学习的西南印度洋深海散射层声学资源密度预测
Prediction on acoustic resource density of deep scattering layer of the southwestern Indian Ocean based on machine learning
摘要
Abstract
Predicting the abundance and distribution of deep scattering layer is important to indicate the distribution of marine protected animals,important fishing grounds,and develop fishery resources into the scattering layer.This study used the Nautical Area Scattering Coefficient(NASC)as the resource density indicator of the scattering layer,and used K-means clustering and SSA-XGBoost model to predict the resource density of the scattering layer based on multiple environmental factors in the southwestern Indian Ocean.The results showed that the accuracy of the model prediction is 80.51%,the precision is 76%,and the recall is 78%.The sample data matches the high-density spatial distribution of the predicted data,and the application effect of the model is good.By predicting the density of the scattering layer in different seasons in 2011,it was found that the center of gravity in the high-density area of the scattering layer moved from southeast to northwest,with the latitude of the center of gravity being the largest in spring and the smallest in winter.The dispersion of the center of gravity in the southeast-northwest direction is greater than that in the northeast-southwest direction.This study can provide a new method for elucidating the distribution and resource variation patterns of scattering layers in larger spaces.关键词
深海散射层/机器学习/声学资源密度/西南印度洋Key words
deep scattering layer/machine learning/acoustic resource density/southwestern Indian Ocean分类
农业科技引用本文复制引用
万树杰,陈新军..基于机器学习的西南印度洋深海散射层声学资源密度预测[J].上海海洋大学学报,2024,33(6):1357-1368,12.基金项目
国家重点研发计划(2019YFD0901401) （2019YFD0901401）
