摘要
对海面风速风向的预测在帆船运动中有着非常重要的意义。为了实现海面风速风向的短期预测和分析,采用时间序列预测法对我国某海域的风速风向数据进行建模仿真,实现一步和多步预测。通过一步预测发现时间序列预测存在延时问题,因此提出将时间序列与卡尔曼滤波融合。预测分析结果表明随着步长的增大预测误差也增大;卡尔曼时间序列融合预测法可以有效地解决时延问题并且提高预测精度。最后将预测结果再通过数据分析得到的结果可以直观有效地提升运动员的比赛水平。
The prediction of wind speed and wind direction of sea surface is very important in sailing.In order to realize the short-term forecast and analysis of sea surface wind speed and wind direction,the time series forecasting method is adopted to model and simulate the wind speed and direction data of a certain sea area in China in this paper.And finally one-step and multistep predictions are achieved.Through one-step prediction,it is found that the time series prediction has the time-delay problem.Therefore,this paper proposes the combination of time series and Kalman filter.The analysis results show that the prediction error increases as the step size increases;Kalman time series fusion prediction method can effectively solve the delay problem and improve the prediction accuracy.Finally,the results of the forecast data analysis can intuitively and effectively improve the level of athletes＇ competition.
作者
屈春晓
任久春
朱谦
QU Chunxiao;REN Jiuchun;ZHU Qian(School of Information Science ancl Technology,Fuclan University,Shanghai 20043)
出处
《微型电脑应用》
2018年第11期113-118,共6页
Microcomputer Applications
关键词
风速预测
风向预测
时间序列
ARIMA模型
卡尔曼预测
帆船竞技
数据分析
Wind speed prediction
Wind direction prediction
Time series
ARIMA model
Ealman prediction
Sailing competition
Data analysis
