摘要
针对惯性平台误差系数呈现小样本和变化趋势多样化的特点,提出改进传统GM(1,1)模型中背景值的构造方式和原始数据算子变换的方法。背景值重构方法可以从结构上解决"以直代曲"引起的误差问题;算子变换方法可以弱化原始数据的随机性并增强其单调性,使得到的改进型GM(1,1)模型数据适用范围更广、模拟预测精度更高。通过实例分析,在两种不同趋势数据的情况下,改进型较传统型模拟预测结果精度都更高,曲线拟合程度更好,更能准确反映惯性平台误差系数的变化趋势。
Aiming at the characteristics of small sample and multiple variation trends of inertial platform error coefficient, an improved GM ( 1, 1 ) model is proposed. The model uses the improved background value and operator transformation of original data. The error caused by the "straight instead of curving" is eliminated by the method of background value reconstruction. Operator transformation method can weaken the randomness of the original data and enhance the monotonicity of it. Therefore, the improved GM ( 1, 1 ) model has a wider scope of application and higher forecast accuracy. Instance analysis result shows that: Compared with the traditional models, the improved model has higher accuracy of forecast and better fitting degree, which is more accurate to reflect the variation trend of inertial platform error coefficient.
出处
《电光与控制》
北大核心
2016年第12期77-80,84,共5页
Electronics Optics & Control
基金
国家自然科学基金(61503392
61403399)
陕西省自然科学基金(2015JQ6213)
关键词
惯性平台
灰色模型
算子变换
背景值
预测
inertial platform
grey model
operator transformation
background value
forecast