摘要
为了提高地下水埋深时间序列的预测精度,本文应用Holt-Winters三参数指数平滑法作为预测模型,使用最小二乘支持向量机对残差序列进行预测。由于核参数和惩罚因子在很大程度上直接影响了最小二乘支持向量机的预测性能,本文选用果蝇优化算法对其参数进行优化选取,该方法不仅能够建立最优的混合预测模型,而且能够很好地捕获地下水埋深序列的非线性特征。选用甘肃民勤县大坝乡城西八社地下水监测站点的数据来验证所建模型的预测性能,实验结果表明与传统的单一预测方法相比,本文所建混合预测模型提高了预测精度。
In order to improve the prediction accuracy of the time series of groundwater depth, Holt-Winters three-parameter exponential smoothing method was used as the prediction model. The least squares support vector machine was also used to predict the residual se- quence. As the prediction performance of least squares support vector machine, to a large extent, is directly affected by the kernel pa- rameters and penalty factors, in this paper, the fruit fly optimization algorithm is used to optimize the parameters, which can not only establish the optimal hybrid forecasting model, but also can capture the Non-Linear characteristics and the predicted model was verified by the data collected from the groundwater monitoring station in the Minqin County Daba in Gansu Province. The experimental results show that the hybrid prediction model proposed in this paper improves the prediction accuracy compared with the traditional single pre- diction method.
出处
《洛阳理工学院学报（自然科学版）》
2017年第4期79-84,共6页
Journal of Luoyang Institute of Science and Technology：Natural Science Edition
基金
国家自然科学基金项目(41371435)
