摘要
针对传统灰色多变量离散模型存在未考虑各驱动因素的时滞动态变化特征及未利用各驱动因素往期数据的问题,通过引入驱动信息控制项调整系数T_i和作用系数λ_i,构建新的灰色多变量离散时滞DDGMD(1,N)模型,并推导出模型参数估计及时间响应式.采用灰色扩维识别方法对调整系数T_i进行识别,明晰各驱动因素及其滞后参数;采用粒子群算法对作用系数λ_i进行优化求解,反应驱动因素往期数据对于当期系统行为序列的影响.最后以江苏省能源消费量预测为例,验证了所提出的模型适用于具有时滞特征的小样本数据预测的有效性.
Aiming at the problem that traditional grey multi-variable discrete models don＇t consider the time-delay dynamic characteristics and the utilization of previous data of the driving variables, a new grey multi-variable discrete model, abbreviated as DDGMD（1, N）, is constructed. And a model parameter calculation method is discussed by introducing adjustment coefficient T~ and effect coefficient A~ of the driving information control. Then, the adjustment coefficient is identified by using the method of grey extended dimension identification, which clarifies the driving variables and their time-delay parameters. The effect coefficient is determined, which reflectes the influence of the previous data of driving variables to the system behavioral sequence, by using the particle swarm optimization algorithm. Finally, a real application about the forecast of the energy consumption in Jiangsu Province is used to demonstrate the feasibility and practicability of the DDGMD（1, N） model. The validity of the model is proved to be suitable for the prediction of small sample data with time-delay characteristics.
出处
《控制与决策》
EI
CSCD
北大核心
2017年第9期1672-1680,共9页
Control and Decision
基金
国家自然科学基金项目(71071077
71371098)
中央高校基本科研业务费专项基金项目(NC2012001
NR2013033)
江苏高校哲学社会科学重点研究基地项目(2012JDXM005)
