摘要
通过构造虚拟现实仿真环境 ,利用仿真模型对系统的直观描述性 ,对系统的运行状况进行模拟 ,并为决策者提供了一个观察系统运行的虚拟空间 ,决策者可以“进入”到系统内部 ,根据观察到的系统状况及个人的偏好 ,在虚拟环境中进行一系列的决策。由于现实系统中 ,系统状态转移概率以及报酬函数往往不能显式得到 ,所以传统方法无法求解序贯决策问题。作为一种新的思路 ,利用仿真优化算法 ,根据仿真运行提供的定量信息 ,利用虚拟现实的环境 ,实现序贯决策过程的优化 ,最终可以得到最优或满意的决策策略。
By constructing virtual reality simulation environment and utilizing the intuitive description of simulation to system implementation, the virtual space is provided to decision - maker to observe the dynamic state of system and 'enter' it to make series of decisions according to the system status and his preference. Because the state transferring probability and income function of system can not be obtained in a close form in real system, traditional methods have difficulties in solving such sequential decision problem. However, by using simulation optimization algorithm as a new method, an optimized or satisfied decision strategy can be obtained.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2001年第10期88-92,共5页
Systems Engineering and Electronics
基金
国家自然科学基金重点项目资助课题 ( 79930 90 0 )
关键词
仿真
序贯分析
决策
虚拟现实
Simulation
Sequential analysis
Optimization
Decision making
