摘要
通过小波分解,将油料消耗量的时间序列分解为低频趋势序列和高频细节序列2部分,低频小波系数采用指数平滑法进行预测,高频小波系数采用马尔科夫预测模型进行预测。将低频和高频的预测结果进行小波重构,得到原始时间序列的预测值,为部队油料保障提供一定的参考依据。
The time series of fuel consumption are decomposed into appcoef sequences and detcoef sequences by means of wavelet decomposition. Exponential smoothing method is used to forecast shredded appcoef sequences while Markov model is used to forecast shredded detcoef sequences. The predicted value of original time series is gained through wavelet recon- struction of appcoef sequences and detcoef sequences predicted values. This method provides references for army＇ s fuel supply.
出处
《军事交通学院学报》
2014年第11期86-90,共5页
Journal of Military Transportation University
关键词
小波分析
油料消耗
指数平滑法
马尔科夫模型
wavelet analysis
fuel consumption
exponential smoothing method
Markov model
