沙漠与绿洲气象2018，Vol.12Issue(2)：89-94,6.DOI:10.12057/j.issn.1002-0799.2018.02.014
两种不同模型对漂流适宜性等级预报的对比分析
Comparison and Analysis of Two Different Models for Drifting Suitability Grade Prediction
摘要
Abstract
In this paper the 2-year daily rafting tourist amount and the previous day's K meteorological elements of Taoyuan River scenic area from January 1st,2015 to December 31st, 2016 were analyzed by using conventional ground meteorological observation data.Both BP network and discriminant analysis methods have good nonlinear function fitting effects,and they were separately used and tested for establishing model.As a meteorological factor for calculating rafting grade prediction,combined with daily experience,the main affecting weather factors were selected (daily precipitation,highest temperature,lowest temperature,average temperature,average relative humidity,minimum relative humidity,08-20 precipitation and wind speed)for establishing the grade equation of rafting suitability.Meanwhile,secondary modelling was carried out in order to reduce the impact on the prediction effect from summer holidays,weekends and other statutory holidays.The result shows that BP network prediction model is better,for its prediction grade accuracy is higher than 40%,and the absolute value of 1 accuracy is higher than 70%,it is more accurate and stable than the discriminant analysis method.In addition,the summer holidays has a positive influence on rafting grade forecast.关键词
漂流适宜性等级/BP神经网络/判别分析Key words
drifting suitability grade/BP netword/discriminant analysis分类
管理科学引用本文复制引用
尚媛媛,杜正静,夏晓玲..两种不同模型对漂流适宜性等级预报的对比分析[J].沙漠与绿洲气象,2018,12(2):89-94,6.基金项目
专业气象服务(旅游气象)项目"气象条件对漂流适宜度影响分析——以桃源河为例"资助. （旅游气象）
