摘要
目前对于大坝变形的预测都是基于单一的预测模型,该模型无法同时考虑样本数据量及影响因子优选情况,而灰色模型适用于短期的预测,并能够很好的确定影响因子,而支持向量机适用于长期的预测,无法确定最佳的影响因子。本文将支持向量机和灰色模型进行很好的结合,可以达到模型互补的目的,提高影响因子的优化及训练样本的拟合和预测样本的预测精度。
The current dam deformation forecast, based on the single forecast model, is unable to consider the condition of the number of sample data and the optimization of influencing factors. The Gray Model （GM） is appli-cable to short-term data and is easy to determine influencing factors. The support vector machine （SVM） is applica-ble to the forecast of long-term data, but cannot determine the best influencing factors. The combination of SVM and GM can achieve the objective of complementary model, and optimize influencing factors and precision of fit-ting and forecasting.
出处
《科技广场》
2015年第3期71-75,共5页
Science Mosaic
