人工智能技术的热带气旋预报综述(之二)——流形学习、智能计算及深度学习的热带气旋预报方法  

Summary of tropical cyclone forecasting based on artificial intelligence technology(part 2)——tropical cyclone forecasting methods based on manifold learning, intelligent calculation and deep learning

作　　者：金龙[1] 黄颖 姚才 黄小燕 赵华生 Jin Long;Huang Ying;Yao Cai;Huang Xiaoyan;Zhao Huasheng(Guangxi Climate Center,Nanning Guangxi 530022;Guangxi Institute of Meteorological Sciences,Nanning Guangxi 530022)

机构地区：[1]广西壮族自治区气候中心,南宁530022 [2]广西壮族自治区气象科学研究所,南宁530022

出　　处：《气象研究与应用》2020年第4期5-12,共8页Journal of Meteorological Research and Application

基　　金：广西自然科学基金(2018GXNSFAA281229,2017GXNSFDA198030,2018GXNSFAA294128);国家自然科学基金(42065004,41765002)。

摘　　要：继"人工智能技术的热带气旋预报综述(之一)"有关BP神经网络和集成方法的热带气旋预报研究和业务应用进行详细综述后,本文将进一步综述流形学习方法在热带气旋预报因子数据挖掘中的应用,以及各种智能计算模型,包括粒子群算法、模糊算法、概率算法,及深度学习方法在热带气旋预报中的应用研究成果,并对今后气象领域的人工智能发展进行初步设想探讨,以期能为有效提高人工智能方法的气象灾害预报能力提供有益参考。Following a detailed review of tropical cyclone forecasting research and operational applications of BP neural network and integrated methods in "Review of Tropical Cyclone Forecast Based on Artificial Intelligence Technology(Part 1)", this article further reviewed the application of manifold learning methods in data mining of tropical cyclone predictors. Besides, the applications of various intelligent computing models,including particle swarm algorithm, fuzzy algorithm, probability algorithm, and deep learning methods in tropical cyclone forecasting were also discussed. A preliminary discussion on the development of artificial intelligence in the field of meteorology in the future is expected to provide a useful reference for effectively improving the ability of artificial intelligence to forecast meteorological disasters.

关 键 词：热带气旋 智能计算 流形学习 数据挖掘 深度学习 

分 类 号：P456[天文地球—大气科学及气象学]