论文部分内容阅读
针对导弹备件消耗呈现“小样本、非平稳”的特点,为了克服传统预测方法依靠大样本数据进行建模的不足,提出了把基于小波变换和改进GM-ARMA的组合预测方法应用于导弹备件消耗预测的构想.在利用小波分解和其他模型建立组合模型的过程中,提出了先对小波基方程和分解层数2个特征进行参数化,再定量地对所有子模型的特征参数进行统一、综合的评估,以达到建立最佳组合模型的目的;然后对具有平稳特性的高频信息用阻尼最小二乘法优化的ARMA(Autoregressive and Moving Average)模型进行预测,对反映整体趋势体现非平稳的低频信息用背景值优化和数据变换技术改进的GM(1,1)模型进行预测.实例结果表明所提出的组合预测方法大大降低了预测误差,说明了该方法的有效性、可行性和实用性.
In order to overcome the shortcomings of the traditional forecasting methods relying on large sample data for modeling the “small sample, non-stationary” consumption of missile spare parts, a new method based on wavelet transform and improved GM-ARMA combined forecasting method is proposed. In the process of using the wavelet decomposition and other models to establish a combined model, two parameters of wavelet basis equation and decomposition layer are firstly parameterized, and the characteristic parameters of all sub-models are re-quantified , Comprehensive assessment to achieve the purpose of establishing the best combination of models; and then forecast the high frequency information with smooth characteristics using damped least square ARMA (Autoregressive and Moving Average) model to predict the overall trend reflects the non-stationary The low frequency information is predicted by GM (1,1) model which is improved by the background value optimization and data transformation technique.Example results show that the proposed combined forecasting method can greatly reduce the prediction error, and illustrates the effectiveness, feasibility and practicability of the proposed method .