以灰色GM(1,1)模型初始条件的优化为主要研究目的,分别讨论了以x^(1)(1)、x^(1)(n)为初始条件的GM(1,1)建模机理;提出了同时利用x^(1)(1)和x^(1)(n)模型初始条件的新方法,以x(1)(t)的模拟值与原始数据的1-AGO序列的误差平方和最小为约束条件,推导了初始条件优化后的预测模型公式;对模型预测精度起决定性作用的参数进行分析,给出了初始条件优化后模型的适用范围。经大量的数据模拟和预测,发现初始条件优化后的GM(1,1)模型各项精度指标均优于传统的GM(1,1)模型。
In order to reestablish the initial value for GM (1,1) model, we proposed a new approach to improve prediction accuracy of GM( 1,1 ) model through optimization of the initial value, which is comprised of the first and the n-th vector as the initialization, and derived from a method of least error summation of square. Then we discussed the parameter which affects the fitting results. By contrasting the improved one to the GM ( 1,1 ) about the simulation and prediction, we can conclude that the improved one is superior in prediction and simulation, and it is proved that the optimum one widen its suitable range.