摘要
复杂系统特征分析是进行系统需求预测的基础和前提。本文将混沌理论应用于复杂预测问题,重点研究了预测系统的相空间重构、关联维数的确定和柯尔莫哥洛夫熵的计算;其次,采用我国1962-2004年大中型拖拉机农具需求量的实测数据进行实证分析,结果表明:我国大中型拖拉机农具需求为混沌系统,影响系统发展预测的因素有7个,最大可预测的时效为6年。
The system forecasting is based on property analysis of complex system. Firstly, a model of the property analysis and forecasting time-scale based on chaos theory was constructed. Then, the paper focused on phase space reconstruction, correlation and kolmogorov-sinaientropy of chaos theory. Secondly, the actual demanding data of big-medium-sized tractor implement from 1962- 2004 was applied to validate the model proposed. The result showed that demand of big-medium-sized tractor implement was a chaos system, which indicated that there should have about 7 independent variables for describing demand change characteristics and the forecasting time-scale is about 6 years.
出处
《微计算机信息》
北大核心
2007年第04X期216-218,共3页
Control & Automation
基金
国家"863"高技术计划项目(2002AA2Z405)
西北农林科技大学人才专项基金(04ZR015)
关键词
复杂预测系统
混沌理论
特征分析
预测时效
Complex system, Chaos theory, Property analysis, Forecasting time-scale
