优化离散灰色幂模型及其应用

发布时间：2018-07-06 11:03

本文选题：灰色系统 + 离散灰色幂模型　；参考：《中国管理科学》2016年02期

【摘要】：考虑已有的灰色预测模型主要能对指数型发展系统或幂函数型发展系统进行模拟预测,本文构建了一种不仅能够模拟指数型和幂函数型的发展系统,并且能够体现出二者之间的相互作用关系的离散灰色幂模型;并针对初始条件对离散灰色幂模型模拟精度的影响,首先给出了离散灰色幂模型的建模步骤,然后以平均相对误差最小化为目标、参数之间的关系为约束条件,构建了离散灰色幂模型初始条件的优化模型,实现对离散灰色幂模型初始条件的优化。结果表明,优化的离散灰色幂模型使得平均相对误差在理论上达到了最小化,其模拟精度和预测精度都高于传统模型。最后,通过中国网络购物人数数据预测和仿真数据分析,说明了本文优化方法的有效性和适用性。
[Abstract]:Considering that the existing grey forecasting model can simulate the exponential development system or power function development system, this paper constructs a development system which can not only simulate exponential and power function type, but also can reflect the dispersion grey power model of the interaction relationship between the two. In this paper, the modeling steps of discrete grey power model are given first. Then, with the objective of minimizing the average relative error and the relation between the parameters as the constraint conditions, the optimization model of the initial condition of the discrete grey power model is constructed, and the optimization of the initial conditions of the discrete grey power model is realized. The results show that the grey power model is optimized. The optimized discrete grey power model makes the average relative error minimized in theory, and its simulation accuracy and prediction accuracy are higher than that of the traditional model. Finally, the effectiveness and applicability of the optimized method are illustrated by the data prediction and simulation data analysis of the Chinese online shopping population.
【作者单位】：江苏师范大学商学院;江苏师范大学计算机学院;
【基金】：国家自然科学基金资助项目(71301064) 教育部人文社科基金项目(12YJC630262)
【分类号】：F224

【参考文献】