摘要
针对武陵山片区天气特点,提出了一种基于Prophet的预测算法,该算法将时间序列数据分解为趋势项、周期性变化项和突出事件项,根据加法原则进行拟合预报和异常检测.实验结果表明,与传统自回归积分滑动平均模型(Autoregressive Integrated Moving Average Model,ARIMA)时序预测算法相比,Prophet算法在时序性、周期性数据的中长期预测中具有明显优势.基于Prophet算法在武陵山片区天气数据的预报和异常检测研究,不仅能为地方政府及时准确地进行天气预报提供技术支撑,还有助于当地民众提前安排生产、生活,减少不必要的损失.
A prediction algorithm based on prophet is proposed for the weather time series data in Wuling mountain area. The algorithm decomposes the time series data into trend items,periodic change items and prominent event items,and carries out fitting prediction and anomaly detection according to the addition principle. According to the experimental results,compared with the traditional autoregressive integrated moving average model(ARIMA) time series prediction algorithm,prophet algorithm has significant advantages in medium and long-term prediction,and the longer the time is,the more obvious the advantage is. The research shows that the prediction and anomaly detection of weather data in Wuling Mountain Area by prophet algorithm not only helps the local people make living and production arrangements in advance and reduce unnecessary losses,but also provides technical support for the local government to make timely and accurate weather forecast.
作者
李森林
印东
LI Sen-lin;YIN Dong(School of Computer Science and Engineering,Huaihua University,Huaihua,Hunan 418008;Key Laboratory of Intelligent Control Technology for Wuling-Mountain Ecological Agriculture in Hunan Province,Huaihua,Hunan 418008)
出处
《怀化学院学报》
2022年第5期54-58,共5页
Journal of Huaihua University
基金
武陵山片区生态农业智能技术湖南省重点实验项目“武陵山片区农业大数据机器学习分析方法与深度神经网络学习模型研究”(ZNKZ2018-6)
湖南省教育厅科学研究一般项目“基于深度学习多类型输入武陵山片区农业产值数据预测模型研究”(19C1483)。
