决策网络计划中决策单元结构及优化方法拓展

为解决决策网络计划中随机样本空间变化而产生的模型表达上的困难,提出经拓展能够描述多个样本空间的决策单元结构.将随机规划理论引入决策网络计划的优化中,建立了新的考虑期望成本与风险等综合因素的数学模型,并通过算例实现了对模型的求解.计算结果表明:经拓展后的决策单元结构及相应的优化模型能够更为有效地解决不同样本空间下的决策问题,具有较高的理论意义与实用价值.

Expandation both of decision unit structure and its optimization method in decision network planning

To solve the expression difficulties resulting from variety of random sample spaces in decision network planning,an expanded decision unit structure including several sample spaces was presented.Stochastic programming theory was used in the optimization of decision network planning and a new mathematical model considering both of expected cost and risk was established.A solution was given through an example to validate the present model.The calculation result shows that the expanded decision unit structure and the corresponding optimization model can solve the decision problem more effectively in variable sample spaces,and therefore has theoretical significance and practical value.