摘要
目前通过数学仿真模型对数据进行分析和预测是很常用的手段,但是现有模型计算平台的实现都是针对某一特定模型,并没有一种统一的模型描述、计算架构。为解决上述问题,提出了一种针对线性分析预测模型计算平台的通用解决方案。研究了该模型中涉及到的数据形式以及数据间的关系,该方案定义了包含模型描述、模型计算、数据提供的三层结构,并且针对部分关键组件的实现方式进行了详细描述。
Using mathematical simulation model to analyze and make a forecast is very normal nowadays. However, amodel-computing program is always designed for a special model, so there is no such a unified computing platform that most models can be work on it. To solve these questions, one unified solution, which aims at linear analyze and forecast models computing platform, is put forward. According to the research on the data structures of model and the rules in them, this solution defines the three-layer architecture that contains description, computing and data providing, also describes the realization of the key components in detail.
出处
《计算机工程与设计》
CSCD
北大核心
2009年第4期977-980,共4页
Computer Engineering and Design
关键词
线性模型
模型计算
分析预测
数据仓库
linear model
model computing
analyze and forecast
data ware
